DL_Final_Project.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "code",
      "execution_count": 1,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "w5L4EDJFt6_F",
        "outputId": "1dbef110-67b3-4589-a31a-6ba0239506ed"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Archive:  GTdb_crop.zip\n",
            "  inflating: data/cropped_faces/s01_01.jpg  \n",
            "  inflating: data/cropped_faces/s01_02.jpg  \n",
            "  inflating: data/cropped_faces/s01_03.jpg  \n",
            "  inflating: data/cropped_faces/s01_04.jpg  \n",
            "  inflating: data/cropped_faces/s01_05.jpg  \n",
            "  inflating: data/cropped_faces/s01_06.jpg  \n",
            "  inflating: data/cropped_faces/s01_07.jpg  \n",
            "  inflating: data/cropped_faces/s01_08.jpg  \n",
            "  inflating: data/cropped_faces/s01_09.jpg  \n",
            "  inflating: data/cropped_faces/s01_10.jpg  \n",
            "  inflating: data/cropped_faces/s01_11.jpg  \n",
            "  inflating: data/cropped_faces/s01_12.jpg  \n",
            "  inflating: data/cropped_faces/s01_13.jpg  \n",
            "  inflating: data/cropped_faces/s01_14.jpg  \n",
            "  inflating: data/cropped_faces/s01_15.jpg  \n",
            "  inflating: data/cropped_faces/s02_01.jpg  \n",
            "  inflating: data/cropped_faces/s02_02.jpg  \n",
            "  inflating: data/cropped_faces/s02_03.jpg  \n",
            "  inflating: data/cropped_faces/s02_04.jpg  \n",
            "  inflating: data/cropped_faces/s02_05.jpg  \n",
            "  inflating: data/cropped_faces/s02_06.jpg  \n",
            "  inflating: data/cropped_faces/s02_07.jpg  \n",
            "  inflating: data/cropped_faces/s02_08.jpg  \n",
            "  inflating: data/cropped_faces/s02_09.jpg  \n",
            "  inflating: data/cropped_faces/s02_10.jpg  \n",
            "  inflating: data/cropped_faces/s02_11.jpg  \n",
            "  inflating: data/cropped_faces/s02_12.jpg  \n",
            "  inflating: data/cropped_faces/s02_13.jpg  \n",
            "  inflating: data/cropped_faces/s02_14.jpg  \n",
            "  inflating: data/cropped_faces/s02_15.jpg  \n",
            "  inflating: data/cropped_faces/s03_01.jpg  \n",
            "  inflating: data/cropped_faces/s03_02.jpg  \n",
            "  inflating: data/cropped_faces/s03_03.jpg  \n",
            "  inflating: data/cropped_faces/s03_04.jpg  \n",
            "  inflating: data/cropped_faces/s03_05.jpg  \n",
            "  inflating: data/cropped_faces/s03_06.jpg  \n",
            "  inflating: data/cropped_faces/s03_07.jpg  \n",
            "  inflating: data/cropped_faces/s03_08.jpg  \n",
            "  inflating: data/cropped_faces/s03_09.jpg  \n",
            "  inflating: data/cropped_faces/s03_10.jpg  \n",
            "  inflating: data/cropped_faces/s03_11.jpg  \n",
            "  inflating: data/cropped_faces/s03_12.jpg  \n",
            "  inflating: data/cropped_faces/s03_13.jpg  \n",
            "  inflating: data/cropped_faces/s03_14.jpg  \n",
            "  inflating: data/cropped_faces/s03_15.jpg  \n",
            "  inflating: data/cropped_faces/s04_01.jpg  \n",
            "  inflating: data/cropped_faces/s04_02.jpg  \n",
            "  inflating: data/cropped_faces/s04_03.jpg  \n",
            "  inflating: data/cropped_faces/s04_04.jpg  \n",
            "  inflating: data/cropped_faces/s04_05.jpg  \n",
            "  inflating: data/cropped_faces/s04_06.jpg  \n",
            "  inflating: data/cropped_faces/s04_07.jpg  \n",
            "  inflating: data/cropped_faces/s04_08.jpg  \n",
            "  inflating: data/cropped_faces/s04_09.jpg  \n",
            "  inflating: data/cropped_faces/s04_10.jpg  \n",
            "  inflating: data/cropped_faces/s04_11.jpg  \n",
            "  inflating: data/cropped_faces/s04_12.jpg  \n",
            "  inflating: data/cropped_faces/s04_13.jpg  \n",
            "  inflating: data/cropped_faces/s04_14.jpg  \n",
            "  inflating: data/cropped_faces/s04_15.jpg  \n",
            "  inflating: data/cropped_faces/s05_01.jpg  \n",
            "  inflating: data/cropped_faces/s05_02.jpg  \n",
            "  inflating: data/cropped_faces/s05_03.jpg  \n",
            "  inflating: data/cropped_faces/s05_04.jpg  \n",
            "  inflating: data/cropped_faces/s05_05.jpg  \n",
            "  inflating: data/cropped_faces/s05_06.jpg  \n",
            "  inflating: data/cropped_faces/s05_07.jpg  \n",
            "  inflating: data/cropped_faces/s05_08.jpg  \n",
            "  inflating: data/cropped_faces/s05_09.jpg  \n",
            "  inflating: data/cropped_faces/s05_10.jpg  \n",
            "  inflating: data/cropped_faces/s05_11.jpg  \n",
            "  inflating: data/cropped_faces/s05_12.jpg  \n",
            "  inflating: data/cropped_faces/s05_13.jpg  \n",
            "  inflating: data/cropped_faces/s05_14.jpg  \n",
            "  inflating: data/cropped_faces/s05_15.jpg  \n",
            "  inflating: data/cropped_faces/s06_01.jpg  \n",
            "  inflating: data/cropped_faces/s06_02.jpg  \n",
            "  inflating: data/cropped_faces/s06_03.jpg  \n",
            "  inflating: data/cropped_faces/s06_04.jpg  \n",
            "  inflating: data/cropped_faces/s06_05.jpg  \n",
            "  inflating: data/cropped_faces/s06_06.jpg  \n",
            "  inflating: data/cropped_faces/s06_07.jpg  \n",
            "  inflating: data/cropped_faces/s06_08.jpg  \n",
            "  inflating: data/cropped_faces/s06_09.jpg  \n",
            "  inflating: data/cropped_faces/s06_10.jpg  \n",
            "  inflating: data/cropped_faces/s06_11.jpg  \n",
            "  inflating: data/cropped_faces/s06_12.jpg  \n",
            "  inflating: data/cropped_faces/s06_13.jpg  \n",
            "  inflating: data/cropped_faces/s06_14.jpg  \n",
            "  inflating: data/cropped_faces/s06_15.jpg  \n",
            "  inflating: data/cropped_faces/s07_01.jpg  \n",
            "  inflating: data/cropped_faces/s07_02.jpg  \n",
            "  inflating: data/cropped_faces/s07_03.jpg  \n",
            "  inflating: data/cropped_faces/s07_04.jpg  \n",
            "  inflating: data/cropped_faces/s07_05.jpg  \n",
            "  inflating: data/cropped_faces/s07_06.jpg  \n",
            "  inflating: data/cropped_faces/s07_07.jpg  \n",
            "  inflating: data/cropped_faces/s07_08.jpg  \n",
            "  inflating: data/cropped_faces/s07_09.jpg  \n",
            "  inflating: data/cropped_faces/s07_10.jpg  \n",
            "  inflating: data/cropped_faces/s07_11.jpg  \n",
            "  inflating: data/cropped_faces/s07_12.jpg  \n",
            "  inflating: data/cropped_faces/s07_13.jpg  \n",
            "  inflating: data/cropped_faces/s07_14.jpg  \n",
            "  inflating: data/cropped_faces/s07_15.jpg  \n",
            "  inflating: data/cropped_faces/s08_01.jpg  \n",
            "  inflating: data/cropped_faces/s08_02.jpg  \n",
            "  inflating: data/cropped_faces/s08_03.jpg  \n",
            "  inflating: data/cropped_faces/s08_04.jpg  \n",
            "  inflating: data/cropped_faces/s08_05.jpg  \n",
            "  inflating: data/cropped_faces/s08_06.jpg  \n",
            "  inflating: data/cropped_faces/s08_07.jpg  \n",
            "  inflating: data/cropped_faces/s08_08.jpg  \n",
            "  inflating: data/cropped_faces/s08_09.jpg  \n",
            "  inflating: data/cropped_faces/s08_10.jpg  \n",
            "  inflating: data/cropped_faces/s08_11.jpg  \n",
            "  inflating: data/cropped_faces/s08_12.jpg  \n",
            "  inflating: data/cropped_faces/s08_13.jpg  \n",
            "  inflating: data/cropped_faces/s08_14.jpg  \n",
            "  inflating: data/cropped_faces/s08_15.jpg  \n",
            "  inflating: data/cropped_faces/s09_01.jpg  \n",
            "  inflating: data/cropped_faces/s09_02.jpg  \n",
            "  inflating: data/cropped_faces/s09_03.jpg  \n",
            "  inflating: data/cropped_faces/s09_04.jpg  \n",
            "  inflating: data/cropped_faces/s09_05.jpg  \n",
            "  inflating: data/cropped_faces/s09_06.jpg  \n",
            "  inflating: data/cropped_faces/s09_07.jpg  \n",
            "  inflating: data/cropped_faces/s09_08.jpg  \n",
            "  inflating: data/cropped_faces/s09_09.jpg  \n",
            "  inflating: data/cropped_faces/s09_10.jpg  \n",
            "  inflating: data/cropped_faces/s09_11.jpg  \n",
            "  inflating: data/cropped_faces/s09_12.jpg  \n",
            "  inflating: data/cropped_faces/s09_13.jpg  \n",
            "  inflating: data/cropped_faces/s09_14.jpg  \n",
            "  inflating: data/cropped_faces/s09_15.jpg  \n",
            "  inflating: data/cropped_faces/s10_01.jpg  \n",
            "  inflating: data/cropped_faces/s10_02.jpg  \n",
            "  inflating: data/cropped_faces/s10_03.jpg  \n",
            "  inflating: data/cropped_faces/s10_04.jpg  \n",
            "  inflating: data/cropped_faces/s10_05.jpg  \n",
            "  inflating: data/cropped_faces/s10_06.jpg  \n",
            "  inflating: data/cropped_faces/s10_07.jpg  \n",
            "  inflating: data/cropped_faces/s10_08.jpg  \n",
            "  inflating: data/cropped_faces/s10_09.jpg  \n",
            "  inflating: data/cropped_faces/s10_10.jpg  \n",
            "  inflating: data/cropped_faces/s10_11.jpg  \n",
            "  inflating: data/cropped_faces/s10_12.jpg  \n",
            "  inflating: data/cropped_faces/s10_13.jpg  \n",
            "  inflating: data/cropped_faces/s10_14.jpg  \n",
            "  inflating: data/cropped_faces/s10_15.jpg  \n",
            "  inflating: data/cropped_faces/s11_01.jpg  \n",
            "  inflating: data/cropped_faces/s11_02.jpg  \n",
            "  inflating: data/cropped_faces/s11_03.jpg  \n",
            "  inflating: data/cropped_faces/s11_04.jpg  \n",
            "  inflating: data/cropped_faces/s11_05.jpg  \n",
            "  inflating: data/cropped_faces/s11_06.jpg  \n",
            "  inflating: data/cropped_faces/s11_07.jpg  \n",
            "  inflating: data/cropped_faces/s11_08.jpg  \n",
            "  inflating: data/cropped_faces/s11_09.jpg  \n",
            "  inflating: data/cropped_faces/s11_10.jpg  \n",
            "  inflating: data/cropped_faces/s11_11.jpg  \n",
            "  inflating: data/cropped_faces/s11_12.jpg  \n",
            "  inflating: data/cropped_faces/s11_13.jpg  \n",
            "  inflating: data/cropped_faces/s11_14.jpg  \n",
            "  inflating: data/cropped_faces/s11_15.jpg  \n",
            "  inflating: data/cropped_faces/s12_01.jpg  \n",
            "  inflating: data/cropped_faces/s12_02.jpg  \n",
            "  inflating: data/cropped_faces/s12_03.jpg  \n",
            "  inflating: data/cropped_faces/s12_04.jpg  \n",
            "  inflating: data/cropped_faces/s12_05.jpg  \n",
            "  inflating: data/cropped_faces/s12_06.jpg  \n",
            "  inflating: data/cropped_faces/s12_07.jpg  \n",
            "  inflating: data/cropped_faces/s12_08.jpg  \n",
            "  inflating: data/cropped_faces/s12_09.jpg  \n",
            "  inflating: data/cropped_faces/s12_10.jpg  \n",
            "  inflating: data/cropped_faces/s12_11.jpg  \n",
            "  inflating: data/cropped_faces/s12_12.jpg  \n",
            "  inflating: data/cropped_faces/s12_13.jpg  \n",
            "  inflating: data/cropped_faces/s12_14.jpg  \n",
            "  inflating: data/cropped_faces/s12_15.jpg  \n",
            "  inflating: data/cropped_faces/s13_01.jpg  \n",
            "  inflating: data/cropped_faces/s13_02.jpg  \n",
            "  inflating: data/cropped_faces/s13_03.jpg  \n",
            "  inflating: data/cropped_faces/s13_04.jpg  \n",
            "  inflating: data/cropped_faces/s13_05.jpg  \n",
            "  inflating: data/cropped_faces/s13_06.jpg  \n",
            "  inflating: data/cropped_faces/s13_07.jpg  \n",
            "  inflating: data/cropped_faces/s13_08.jpg  \n",
            "  inflating: data/cropped_faces/s13_09.jpg  \n",
            "  inflating: data/cropped_faces/s13_10.jpg  \n",
            "  inflating: data/cropped_faces/s13_11.jpg  \n",
            "  inflating: data/cropped_faces/s13_12.jpg  \n",
            "  inflating: data/cropped_faces/s13_13.jpg  \n",
            "  inflating: data/cropped_faces/s13_14.jpg  \n",
            "  inflating: data/cropped_faces/s13_15.jpg  \n",
            "  inflating: data/cropped_faces/s14_01.jpg  \n",
            "  inflating: data/cropped_faces/s14_02.jpg  \n",
            "  inflating: data/cropped_faces/s14_03.jpg  \n",
            "  inflating: data/cropped_faces/s14_04.jpg  \n",
            "  inflating: data/cropped_faces/s14_05.jpg  \n",
            "  inflating: data/cropped_faces/s14_06.jpg  \n",
            "  inflating: data/cropped_faces/s14_07.jpg  \n",
            "  inflating: data/cropped_faces/s14_08.jpg  \n",
            "  inflating: data/cropped_faces/s14_09.jpg  \n",
            "  inflating: data/cropped_faces/s14_10.jpg  \n",
            "  inflating: data/cropped_faces/s14_11.jpg  \n",
            "  inflating: data/cropped_faces/s14_12.jpg  \n",
            "  inflating: data/cropped_faces/s14_13.jpg  \n",
            "  inflating: data/cropped_faces/s14_14.jpg  \n",
            "  inflating: data/cropped_faces/s14_15.jpg  \n",
            "  inflating: data/cropped_faces/s15_01.jpg  \n",
            "  inflating: data/cropped_faces/s15_02.jpg  \n",
            "  inflating: data/cropped_faces/s15_03.jpg  \n",
            "  inflating: data/cropped_faces/s15_04.jpg  \n",
            "  inflating: data/cropped_faces/s15_05.jpg  \n",
            "  inflating: data/cropped_faces/s15_06.jpg  \n",
            "  inflating: data/cropped_faces/s15_07.jpg  \n",
            "  inflating: data/cropped_faces/s15_08.jpg  \n",
            "  inflating: data/cropped_faces/s15_09.jpg  \n",
            "  inflating: data/cropped_faces/s15_10.jpg  \n",
            "  inflating: data/cropped_faces/s15_11.jpg  \n",
            "  inflating: data/cropped_faces/s15_12.jpg  \n",
            "  inflating: data/cropped_faces/s15_13.jpg  \n",
            "  inflating: data/cropped_faces/s15_14.jpg  \n",
            "  inflating: data/cropped_faces/s15_15.jpg  \n",
            "  inflating: data/cropped_faces/s16_01.jpg  \n",
            "  inflating: data/cropped_faces/s16_02.jpg  \n",
            "  inflating: data/cropped_faces/s16_03.jpg  \n",
            "  inflating: data/cropped_faces/s16_04.jpg  \n",
            "  inflating: data/cropped_faces/s16_05.jpg  \n",
            "  inflating: data/cropped_faces/s16_06.jpg  \n",
            "  inflating: data/cropped_faces/s16_07.jpg  \n",
            "  inflating: data/cropped_faces/s16_08.jpg  \n",
            "  inflating: data/cropped_faces/s16_09.jpg  \n",
            "  inflating: data/cropped_faces/s16_10.jpg  \n",
            "  inflating: data/cropped_faces/s16_11.jpg  \n",
            "  inflating: data/cropped_faces/s16_12.jpg  \n",
            "  inflating: data/cropped_faces/s16_13.jpg  \n",
            "  inflating: data/cropped_faces/s16_14.jpg  \n",
            "  inflating: data/cropped_faces/s16_15.jpg  \n",
            "  inflating: data/cropped_faces/s17_01.jpg  \n",
            "  inflating: data/cropped_faces/s17_02.jpg  \n",
            "  inflating: data/cropped_faces/s17_03.jpg  \n",
            "  inflating: data/cropped_faces/s17_04.jpg  \n",
            "  inflating: data/cropped_faces/s17_05.jpg  \n",
            "  inflating: data/cropped_faces/s17_06.jpg  \n",
            "  inflating: data/cropped_faces/s17_07.jpg  \n",
            "  inflating: data/cropped_faces/s17_08.jpg  \n",
            "  inflating: data/cropped_faces/s17_09.jpg  \n",
            "  inflating: data/cropped_faces/s17_10.jpg  \n",
            "  inflating: data/cropped_faces/s17_11.jpg  \n",
            "  inflating: data/cropped_faces/s17_12.jpg  \n",
            "  inflating: data/cropped_faces/s17_13.jpg  \n",
            "  inflating: data/cropped_faces/s17_14.jpg  \n",
            "  inflating: data/cropped_faces/s17_15.jpg  \n",
            "  inflating: data/cropped_faces/s18_01.jpg  \n",
            "  inflating: data/cropped_faces/s18_02.jpg  \n",
            "  inflating: data/cropped_faces/s18_03.jpg  \n",
            "  inflating: data/cropped_faces/s18_04.jpg  \n",
            "  inflating: data/cropped_faces/s18_05.jpg  \n",
            "  inflating: data/cropped_faces/s18_06.jpg  \n",
            "  inflating: data/cropped_faces/s18_07.jpg  \n",
            "  inflating: data/cropped_faces/s18_08.jpg  \n",
            "  inflating: data/cropped_faces/s18_09.jpg  \n",
            "  inflating: data/cropped_faces/s18_10.jpg  \n",
            "  inflating: data/cropped_faces/s18_11.jpg  \n",
            "  inflating: data/cropped_faces/s18_12.jpg  \n",
            "  inflating: data/cropped_faces/s18_13.jpg  \n",
            "  inflating: data/cropped_faces/s18_14.jpg  \n",
            "  inflating: data/cropped_faces/s18_15.jpg  \n",
            "  inflating: data/cropped_faces/s19_01.jpg  \n",
            "  inflating: data/cropped_faces/s19_02.jpg  \n",
            "  inflating: data/cropped_faces/s19_03.jpg  \n",
            "  inflating: data/cropped_faces/s19_04.jpg  \n",
            "  inflating: data/cropped_faces/s19_05.jpg  \n",
            "  inflating: data/cropped_faces/s19_06.jpg  \n",
            "  inflating: data/cropped_faces/s19_07.jpg  \n",
            "  inflating: data/cropped_faces/s19_08.jpg  \n",
            "  inflating: data/cropped_faces/s19_09.jpg  \n",
            "  inflating: data/cropped_faces/s19_10.jpg  \n",
            "  inflating: data/cropped_faces/s19_11.jpg  \n",
            "  inflating: data/cropped_faces/s19_12.jpg  \n",
            "  inflating: data/cropped_faces/s19_13.jpg  \n",
            "  inflating: data/cropped_faces/s19_14.jpg  \n",
            "  inflating: data/cropped_faces/s19_15.jpg  \n",
            "  inflating: data/cropped_faces/s20_01.jpg  \n",
            "  inflating: data/cropped_faces/s20_02.jpg  \n",
            "  inflating: data/cropped_faces/s20_03.jpg  \n",
            "  inflating: data/cropped_faces/s20_04.jpg  \n",
            "  inflating: data/cropped_faces/s20_05.jpg  \n",
            "  inflating: data/cropped_faces/s20_06.jpg  \n",
            "  inflating: data/cropped_faces/s20_07.jpg  \n",
            "  inflating: data/cropped_faces/s20_08.jpg  \n",
            "  inflating: data/cropped_faces/s20_09.jpg  \n",
            "  inflating: data/cropped_faces/s20_10.jpg  \n",
            "  inflating: data/cropped_faces/s20_11.jpg  \n",
            "  inflating: data/cropped_faces/s20_12.jpg  \n",
            "  inflating: data/cropped_faces/s20_13.jpg  \n",
            "  inflating: data/cropped_faces/s20_14.jpg  \n",
            "  inflating: data/cropped_faces/s20_15.jpg  \n",
            "  inflating: data/cropped_faces/s21_01.jpg  \n",
            "  inflating: data/cropped_faces/s21_02.jpg  \n",
            "  inflating: data/cropped_faces/s21_03.jpg  \n",
            "  inflating: data/cropped_faces/s21_04.jpg  \n",
            "  inflating: data/cropped_faces/s21_05.jpg  \n",
            "  inflating: data/cropped_faces/s21_06.jpg  \n",
            "  inflating: data/cropped_faces/s21_07.jpg  \n",
            "  inflating: data/cropped_faces/s21_08.jpg  \n",
            "  inflating: data/cropped_faces/s21_09.jpg  \n",
            "  inflating: data/cropped_faces/s21_10.jpg  \n",
            "  inflating: data/cropped_faces/s21_11.jpg  \n",
            "  inflating: data/cropped_faces/s21_12.jpg  \n",
            "  inflating: data/cropped_faces/s21_13.jpg  \n",
            "  inflating: data/cropped_faces/s21_14.jpg  \n",
            "  inflating: data/cropped_faces/s21_15.jpg  \n",
            "  inflating: data/cropped_faces/s22_01.jpg  \n",
            "  inflating: data/cropped_faces/s22_02.jpg  \n",
            "  inflating: data/cropped_faces/s22_03.jpg  \n",
            "  inflating: data/cropped_faces/s22_04.jpg  \n",
            "  inflating: data/cropped_faces/s22_05.jpg  \n",
            "  inflating: data/cropped_faces/s22_06.jpg  \n",
            "  inflating: data/cropped_faces/s22_07.jpg  \n",
            "  inflating: data/cropped_faces/s22_08.jpg  \n",
            "  inflating: data/cropped_faces/s22_09.jpg  \n",
            "  inflating: data/cropped_faces/s22_10.jpg  \n",
            "  inflating: data/cropped_faces/s22_11.jpg  \n",
            "  inflating: data/cropped_faces/s22_12.jpg  \n",
            "  inflating: data/cropped_faces/s22_13.jpg  \n",
            "  inflating: data/cropped_faces/s22_14.jpg  \n",
            "  inflating: data/cropped_faces/s22_15.jpg  \n",
            "  inflating: data/cropped_faces/s23_01.jpg  \n",
            "  inflating: data/cropped_faces/s23_02.jpg  \n",
            "  inflating: data/cropped_faces/s23_03.jpg  \n",
            "  inflating: data/cropped_faces/s23_04.jpg  \n",
            "  inflating: data/cropped_faces/s23_05.jpg  \n",
            "  inflating: data/cropped_faces/s23_06.jpg  \n",
            "  inflating: data/cropped_faces/s23_07.jpg  \n",
            "  inflating: data/cropped_faces/s23_08.jpg  \n",
            "  inflating: data/cropped_faces/s23_09.jpg  \n",
            "  inflating: data/cropped_faces/s23_10.jpg  \n",
            "  inflating: data/cropped_faces/s23_11.jpg  \n",
            "  inflating: data/cropped_faces/s23_12.jpg  \n",
            "  inflating: data/cropped_faces/s23_13.jpg  \n",
            "  inflating: data/cropped_faces/s23_14.jpg  \n",
            "  inflating: data/cropped_faces/s23_15.jpg  \n",
            "  inflating: data/cropped_faces/s24_01.jpg  \n",
            "  inflating: data/cropped_faces/s24_02.jpg  \n",
            "  inflating: data/cropped_faces/s24_03.jpg  \n",
            "  inflating: data/cropped_faces/s24_04.jpg  \n",
            "  inflating: data/cropped_faces/s24_05.jpg  \n",
            "  inflating: data/cropped_faces/s24_06.jpg  \n",
            "  inflating: data/cropped_faces/s24_07.jpg  \n",
            "  inflating: data/cropped_faces/s24_08.jpg  \n",
            "  inflating: data/cropped_faces/s24_09.jpg  \n",
            "  inflating: data/cropped_faces/s24_10.jpg  \n",
            "  inflating: data/cropped_faces/s24_11.jpg  \n",
            "  inflating: data/cropped_faces/s24_12.jpg  \n",
            "  inflating: data/cropped_faces/s24_13.jpg  \n",
            "  inflating: data/cropped_faces/s24_14.jpg  \n",
            "  inflating: data/cropped_faces/s24_15.jpg  \n",
            "  inflating: data/cropped_faces/s25_01.jpg  \n",
            "  inflating: data/cropped_faces/s25_02.jpg  \n",
            "  inflating: data/cropped_faces/s25_03.jpg  \n",
            "  inflating: data/cropped_faces/s25_04.jpg  \n",
            "  inflating: data/cropped_faces/s25_05.jpg  \n",
            "  inflating: data/cropped_faces/s25_06.jpg  \n",
            "  inflating: data/cropped_faces/s25_07.jpg  \n",
            "  inflating: data/cropped_faces/s25_08.jpg  \n",
            "  inflating: data/cropped_faces/s25_09.jpg  \n",
            "  inflating: data/cropped_faces/s25_10.jpg  \n",
            "  inflating: data/cropped_faces/s25_11.jpg  \n",
            "  inflating: data/cropped_faces/s25_12.jpg  \n",
            "  inflating: data/cropped_faces/s25_13.jpg  \n",
            "  inflating: data/cropped_faces/s25_14.jpg  \n",
            "  inflating: data/cropped_faces/s25_15.jpg  \n",
            "  inflating: data/cropped_faces/s26_01.jpg  \n",
            "  inflating: data/cropped_faces/s26_02.jpg  \n",
            "  inflating: data/cropped_faces/s26_03.jpg  \n",
            "  inflating: data/cropped_faces/s26_04.jpg  \n",
            "  inflating: data/cropped_faces/s26_05.jpg  \n",
            "  inflating: data/cropped_faces/s26_06.jpg  \n",
            "  inflating: data/cropped_faces/s26_07.jpg  \n",
            "  inflating: data/cropped_faces/s26_08.jpg  \n",
            "  inflating: data/cropped_faces/s26_09.jpg  \n",
            "  inflating: data/cropped_faces/s26_10.jpg  \n",
            "  inflating: data/cropped_faces/s26_11.jpg  \n",
            "  inflating: data/cropped_faces/s26_12.jpg  \n",
            "  inflating: data/cropped_faces/s26_13.jpg  \n",
            "  inflating: data/cropped_faces/s26_14.jpg  \n",
            "  inflating: data/cropped_faces/s26_15.jpg  \n",
            "  inflating: data/cropped_faces/s27_01.jpg  \n",
            "  inflating: data/cropped_faces/s27_02.jpg  \n",
            "  inflating: data/cropped_faces/s27_03.jpg  \n",
            "  inflating: data/cropped_faces/s27_04.jpg  \n",
            "  inflating: data/cropped_faces/s27_05.jpg  \n",
            "  inflating: data/cropped_faces/s27_06.jpg  \n",
            "  inflating: data/cropped_faces/s27_07.jpg  \n",
            "  inflating: data/cropped_faces/s27_08.jpg  \n",
            "  inflating: data/cropped_faces/s27_09.jpg  \n",
            "  inflating: data/cropped_faces/s27_10.jpg  \n",
            "  inflating: data/cropped_faces/s27_11.jpg  \n",
            "  inflating: data/cropped_faces/s27_12.jpg  \n",
            "  inflating: data/cropped_faces/s27_13.jpg  \n",
            "  inflating: data/cropped_faces/s27_14.jpg  \n",
            "  inflating: data/cropped_faces/s27_15.jpg  \n",
            "  inflating: data/cropped_faces/s28_01.jpg  \n",
            "  inflating: data/cropped_faces/s28_02.jpg  \n",
            "  inflating: data/cropped_faces/s28_03.jpg  \n",
            "  inflating: data/cropped_faces/s28_04.jpg  \n",
            "  inflating: data/cropped_faces/s28_05.jpg  \n",
            "  inflating: data/cropped_faces/s28_06.jpg  \n",
            "  inflating: data/cropped_faces/s28_07.jpg  \n",
            "  inflating: data/cropped_faces/s28_08.jpg  \n",
            "  inflating: data/cropped_faces/s28_09.jpg  \n",
            "  inflating: data/cropped_faces/s28_10.jpg  \n",
            "  inflating: data/cropped_faces/s28_11.jpg  \n",
            "  inflating: data/cropped_faces/s28_12.jpg  \n",
            "  inflating: data/cropped_faces/s28_13.jpg  \n",
            "  inflating: data/cropped_faces/s28_14.jpg  \n",
            "  inflating: data/cropped_faces/s28_15.jpg  \n",
            "  inflating: data/cropped_faces/s29_01.jpg  \n",
            "  inflating: data/cropped_faces/s29_02.jpg  \n",
            "  inflating: data/cropped_faces/s29_03.jpg  \n",
            "  inflating: data/cropped_faces/s29_04.jpg  \n",
            "  inflating: data/cropped_faces/s29_05.jpg  \n",
            "  inflating: data/cropped_faces/s29_06.jpg  \n",
            "  inflating: data/cropped_faces/s29_07.jpg  \n",
            "  inflating: data/cropped_faces/s29_08.jpg  \n",
            "  inflating: data/cropped_faces/s29_09.jpg  \n",
            "  inflating: data/cropped_faces/s29_10.jpg  \n",
            "  inflating: data/cropped_faces/s29_11.jpg  \n",
            "  inflating: data/cropped_faces/s29_12.jpg  \n",
            "  inflating: data/cropped_faces/s29_13.jpg  \n",
            "  inflating: data/cropped_faces/s29_14.jpg  \n",
            "  inflating: data/cropped_faces/s29_15.jpg  \n",
            "  inflating: data/cropped_faces/s30_01.jpg  \n",
            "  inflating: data/cropped_faces/s30_02.jpg  \n",
            "  inflating: data/cropped_faces/s30_03.jpg  \n",
            "  inflating: data/cropped_faces/s30_04.jpg  \n",
            "  inflating: data/cropped_faces/s30_05.jpg  \n",
            "  inflating: data/cropped_faces/s30_06.jpg  \n",
            "  inflating: data/cropped_faces/s30_07.jpg  \n",
            "  inflating: data/cropped_faces/s30_08.jpg  \n",
            "  inflating: data/cropped_faces/s30_09.jpg  \n",
            "  inflating: data/cropped_faces/s30_10.jpg  \n",
            "  inflating: data/cropped_faces/s30_11.jpg  \n",
            "  inflating: data/cropped_faces/s30_12.jpg  \n",
            "  inflating: data/cropped_faces/s30_13.jpg  \n",
            "  inflating: data/cropped_faces/s30_14.jpg  \n",
            "  inflating: data/cropped_faces/s30_15.jpg  \n",
            "  inflating: data/cropped_faces/s31_01.jpg  \n",
            "  inflating: data/cropped_faces/s31_02.jpg  \n",
            "  inflating: data/cropped_faces/s31_03.jpg  \n",
            "  inflating: data/cropped_faces/s31_04.jpg  \n",
            "  inflating: data/cropped_faces/s31_05.jpg  \n",
            "  inflating: data/cropped_faces/s31_06.jpg  \n",
            "  inflating: data/cropped_faces/s31_07.jpg  \n",
            "  inflating: data/cropped_faces/s31_08.jpg  \n",
            "  inflating: data/cropped_faces/s31_09.jpg  \n",
            "  inflating: data/cropped_faces/s31_10.jpg  \n",
            "  inflating: data/cropped_faces/s31_11.jpg  \n",
            "  inflating: data/cropped_faces/s31_12.jpg  \n",
            "  inflating: data/cropped_faces/s31_13.jpg  \n",
            "  inflating: data/cropped_faces/s31_14.jpg  \n",
            "  inflating: data/cropped_faces/s31_15.jpg  \n",
            "  inflating: data/cropped_faces/s32_01.jpg  \n",
            "  inflating: data/cropped_faces/s32_02.jpg  \n",
            "  inflating: data/cropped_faces/s32_03.jpg  \n",
            "  inflating: data/cropped_faces/s32_04.jpg  \n",
            "  inflating: data/cropped_faces/s32_05.jpg  \n",
            "  inflating: data/cropped_faces/s32_06.jpg  \n",
            "  inflating: data/cropped_faces/s32_07.jpg  \n",
            "  inflating: data/cropped_faces/s32_08.jpg  \n",
            "  inflating: data/cropped_faces/s32_09.jpg  \n",
            "  inflating: data/cropped_faces/s32_10.jpg  \n",
            "  inflating: data/cropped_faces/s32_11.jpg  \n",
            "  inflating: data/cropped_faces/s32_12.jpg  \n",
            "  inflating: data/cropped_faces/s32_13.jpg  \n",
            "  inflating: data/cropped_faces/s32_14.jpg  \n",
            "  inflating: data/cropped_faces/s32_15.jpg  \n",
            "  inflating: data/cropped_faces/s33_01.jpg  \n",
            "  inflating: data/cropped_faces/s33_02.jpg  \n",
            "  inflating: data/cropped_faces/s33_03.jpg  \n",
            "  inflating: data/cropped_faces/s33_04.jpg  \n",
            "  inflating: data/cropped_faces/s33_05.jpg  \n",
            "  inflating: data/cropped_faces/s33_06.jpg  \n",
            "  inflating: data/cropped_faces/s33_07.jpg  \n",
            "  inflating: data/cropped_faces/s33_08.jpg  \n",
            "  inflating: data/cropped_faces/s33_09.jpg  \n",
            "  inflating: data/cropped_faces/s33_10.jpg  \n",
            "  inflating: data/cropped_faces/s33_11.jpg  \n",
            "  inflating: data/cropped_faces/s33_12.jpg  \n",
            "  inflating: data/cropped_faces/s33_13.jpg  \n",
            "  inflating: data/cropped_faces/s33_14.jpg  \n",
            "  inflating: data/cropped_faces/s33_15.jpg  \n",
            "  inflating: data/cropped_faces/s34_01.jpg  \n",
            "  inflating: data/cropped_faces/s34_02.jpg  \n",
            "  inflating: data/cropped_faces/s34_03.jpg  \n",
            "  inflating: data/cropped_faces/s34_04.jpg  \n",
            "  inflating: data/cropped_faces/s34_05.jpg  \n",
            "  inflating: data/cropped_faces/s34_06.jpg  \n",
            "  inflating: data/cropped_faces/s34_07.jpg  \n",
            "  inflating: data/cropped_faces/s34_08.jpg  \n",
            "  inflating: data/cropped_faces/s34_09.jpg  \n",
            "  inflating: data/cropped_faces/s34_10.jpg  \n",
            "  inflating: data/cropped_faces/s34_11.jpg  \n",
            "  inflating: data/cropped_faces/s34_12.jpg  \n",
            "  inflating: data/cropped_faces/s34_13.jpg  \n",
            "  inflating: data/cropped_faces/s34_14.jpg  \n",
            "  inflating: data/cropped_faces/s34_15.jpg  \n",
            "  inflating: data/cropped_faces/s35_01.jpg  \n",
            "  inflating: data/cropped_faces/s35_02.jpg  \n",
            "  inflating: data/cropped_faces/s35_03.jpg  \n",
            "  inflating: data/cropped_faces/s35_04.jpg  \n",
            "  inflating: data/cropped_faces/s35_05.jpg  \n",
            "  inflating: data/cropped_faces/s35_06.jpg  \n",
            "  inflating: data/cropped_faces/s35_07.jpg  \n",
            "  inflating: data/cropped_faces/s35_08.jpg  \n",
            "  inflating: data/cropped_faces/s35_09.jpg  \n",
            "  inflating: data/cropped_faces/s35_10.jpg  \n",
            "  inflating: data/cropped_faces/s35_11.jpg  \n",
            "  inflating: data/cropped_faces/s35_12.jpg  \n",
            "  inflating: data/cropped_faces/s35_13.jpg  \n",
            "  inflating: data/cropped_faces/s35_14.jpg  \n",
            "  inflating: data/cropped_faces/s35_15.jpg  \n",
            "  inflating: data/cropped_faces/s36_01.jpg  \n",
            "  inflating: data/cropped_faces/s36_02.jpg  \n",
            "  inflating: data/cropped_faces/s36_03.jpg  \n",
            "  inflating: data/cropped_faces/s36_04.jpg  \n",
            "  inflating: data/cropped_faces/s36_05.jpg  \n",
            "  inflating: data/cropped_faces/s36_06.jpg  \n",
            "  inflating: data/cropped_faces/s36_07.jpg  \n",
            "  inflating: data/cropped_faces/s36_08.jpg  \n",
            "  inflating: data/cropped_faces/s36_09.jpg  \n",
            "  inflating: data/cropped_faces/s36_10.jpg  \n",
            "  inflating: data/cropped_faces/s36_11.jpg  \n",
            "  inflating: data/cropped_faces/s36_12.jpg  \n",
            "  inflating: data/cropped_faces/s36_13.jpg  \n",
            "  inflating: data/cropped_faces/s36_14.jpg  \n",
            "  inflating: data/cropped_faces/s36_15.jpg  \n",
            "  inflating: data/cropped_faces/s37_01.jpg  \n",
            "  inflating: data/cropped_faces/s37_02.jpg  \n",
            "  inflating: data/cropped_faces/s37_03.jpg  \n",
            "  inflating: data/cropped_faces/s37_04.jpg  \n",
            "  inflating: data/cropped_faces/s37_05.jpg  \n",
            "  inflating: data/cropped_faces/s37_06.jpg  \n",
            "  inflating: data/cropped_faces/s37_07.jpg  \n",
            "  inflating: data/cropped_faces/s37_08.jpg  \n",
            "  inflating: data/cropped_faces/s37_09.jpg  \n",
            "  inflating: data/cropped_faces/s37_10.jpg  \n",
            "  inflating: data/cropped_faces/s37_11.jpg  \n",
            "  inflating: data/cropped_faces/s37_12.jpg  \n",
            "  inflating: data/cropped_faces/s37_13.jpg  \n",
            "  inflating: data/cropped_faces/s37_14.jpg  \n",
            "  inflating: data/cropped_faces/s37_15.jpg  \n",
            "  inflating: data/cropped_faces/s38_01.jpg  \n",
            "  inflating: data/cropped_faces/s38_02.jpg  \n",
            "  inflating: data/cropped_faces/s38_03.jpg  \n",
            "  inflating: data/cropped_faces/s38_04.jpg  \n",
            "  inflating: data/cropped_faces/s38_05.jpg  \n",
            "  inflating: data/cropped_faces/s38_06.jpg  \n",
            "  inflating: data/cropped_faces/s38_07.jpg  \n",
            "  inflating: data/cropped_faces/s38_08.jpg  \n",
            "  inflating: data/cropped_faces/s38_09.jpg  \n",
            "  inflating: data/cropped_faces/s38_10.jpg  \n",
            "  inflating: data/cropped_faces/s38_11.jpg  \n",
            "  inflating: data/cropped_faces/s38_12.jpg  \n",
            "  inflating: data/cropped_faces/s38_13.jpg  \n",
            "  inflating: data/cropped_faces/s38_14.jpg  \n",
            "  inflating: data/cropped_faces/s38_15.jpg  \n",
            "  inflating: data/cropped_faces/s39_01.jpg  \n",
            "  inflating: data/cropped_faces/s39_02.jpg  \n",
            "  inflating: data/cropped_faces/s39_03.jpg  \n",
            "  inflating: data/cropped_faces/s39_04.jpg  \n",
            "  inflating: data/cropped_faces/s39_05.jpg  \n",
            "  inflating: data/cropped_faces/s39_06.jpg  \n",
            "  inflating: data/cropped_faces/s39_07.jpg  \n",
            "  inflating: data/cropped_faces/s39_08.jpg  \n",
            "  inflating: data/cropped_faces/s39_09.jpg  \n",
            "  inflating: data/cropped_faces/s39_10.jpg  \n",
            "  inflating: data/cropped_faces/s39_11.jpg  \n",
            "  inflating: data/cropped_faces/s39_12.jpg  \n",
            "  inflating: data/cropped_faces/s39_13.jpg  \n",
            "  inflating: data/cropped_faces/s39_14.jpg  \n",
            "  inflating: data/cropped_faces/s39_15.jpg  \n",
            "  inflating: data/cropped_faces/s40_01.jpg  \n",
            "  inflating: data/cropped_faces/s40_02.jpg  \n",
            "  inflating: data/cropped_faces/s40_03.jpg  \n",
            "  inflating: data/cropped_faces/s40_04.jpg  \n",
            "  inflating: data/cropped_faces/s40_05.jpg  \n",
            "  inflating: data/cropped_faces/s40_06.jpg  \n",
            "  inflating: data/cropped_faces/s40_07.jpg  \n",
            "  inflating: data/cropped_faces/s40_08.jpg  \n",
            "  inflating: data/cropped_faces/s40_09.jpg  \n",
            "  inflating: data/cropped_faces/s40_10.jpg  \n",
            "  inflating: data/cropped_faces/s40_11.jpg  \n",
            "  inflating: data/cropped_faces/s40_12.jpg  \n",
            "  inflating: data/cropped_faces/s40_13.jpg  \n",
            "  inflating: data/cropped_faces/s40_14.jpg  \n",
            "  inflating: data/cropped_faces/s40_15.jpg  \n",
            "  inflating: data/cropped_faces/s41_01.jpg  \n",
            "  inflating: data/cropped_faces/s41_02.jpg  \n",
            "  inflating: data/cropped_faces/s41_03.jpg  \n",
            "  inflating: data/cropped_faces/s41_04.jpg  \n",
            "  inflating: data/cropped_faces/s41_05.jpg  \n",
            "  inflating: data/cropped_faces/s41_06.jpg  \n",
            "  inflating: data/cropped_faces/s41_07.jpg  \n",
            "  inflating: data/cropped_faces/s41_08.jpg  \n",
            "  inflating: data/cropped_faces/s41_09.jpg  \n",
            "  inflating: data/cropped_faces/s41_10.jpg  \n",
            "  inflating: data/cropped_faces/s41_11.jpg  \n",
            "  inflating: data/cropped_faces/s41_12.jpg  \n",
            "  inflating: data/cropped_faces/s41_13.jpg  \n",
            "  inflating: data/cropped_faces/s41_14.jpg  \n",
            "  inflating: data/cropped_faces/s41_15.jpg  \n",
            "  inflating: data/cropped_faces/s42_01.jpg  \n",
            "  inflating: data/cropped_faces/s42_02.jpg  \n",
            "  inflating: data/cropped_faces/s42_03.jpg  \n",
            "  inflating: data/cropped_faces/s42_04.jpg  \n",
            "  inflating: data/cropped_faces/s42_05.jpg  \n",
            "  inflating: data/cropped_faces/s42_06.jpg  \n",
            "  inflating: data/cropped_faces/s42_07.jpg  \n",
            "  inflating: data/cropped_faces/s42_08.jpg  \n",
            "  inflating: data/cropped_faces/s42_09.jpg  \n",
            "  inflating: data/cropped_faces/s42_10.jpg  \n",
            "  inflating: data/cropped_faces/s42_11.jpg  \n",
            "  inflating: data/cropped_faces/s42_12.jpg  \n",
            "  inflating: data/cropped_faces/s42_13.jpg  \n",
            "  inflating: data/cropped_faces/s42_14.jpg  \n",
            "  inflating: data/cropped_faces/s42_15.jpg  \n",
            "  inflating: data/cropped_faces/s43_01.jpg  \n",
            "  inflating: data/cropped_faces/s43_02.jpg  \n",
            "  inflating: data/cropped_faces/s43_03.jpg  \n",
            "  inflating: data/cropped_faces/s43_04.jpg  \n",
            "  inflating: data/cropped_faces/s43_05.jpg  \n",
            "  inflating: data/cropped_faces/s43_06.jpg  \n",
            "  inflating: data/cropped_faces/s43_07.jpg  \n",
            "  inflating: data/cropped_faces/s43_08.jpg  \n",
            "  inflating: data/cropped_faces/s43_09.jpg  \n",
            "  inflating: data/cropped_faces/s43_10.jpg  \n",
            "  inflating: data/cropped_faces/s43_11.jpg  \n",
            "  inflating: data/cropped_faces/s43_12.jpg  \n",
            "  inflating: data/cropped_faces/s43_13.jpg  \n",
            "  inflating: data/cropped_faces/s43_14.jpg  \n",
            "  inflating: data/cropped_faces/s43_15.jpg  \n",
            "  inflating: data/cropped_faces/s44_01.jpg  \n",
            "  inflating: data/cropped_faces/s44_02.jpg  \n",
            "  inflating: data/cropped_faces/s44_03.jpg  \n",
            "  inflating: data/cropped_faces/s44_04.jpg  \n",
            "  inflating: data/cropped_faces/s44_05.jpg  \n",
            "  inflating: data/cropped_faces/s44_06.jpg  \n",
            "  inflating: data/cropped_faces/s44_07.jpg  \n",
            "  inflating: data/cropped_faces/s44_08.jpg  \n",
            "  inflating: data/cropped_faces/s44_09.jpg  \n",
            "  inflating: data/cropped_faces/s44_10.jpg  \n",
            "  inflating: data/cropped_faces/s44_11.jpg  \n",
            "  inflating: data/cropped_faces/s44_12.jpg  \n",
            "  inflating: data/cropped_faces/s44_13.jpg  \n",
            "  inflating: data/cropped_faces/s44_14.jpg  \n",
            "  inflating: data/cropped_faces/s44_15.jpg  \n",
            "  inflating: data/cropped_faces/s45_01.jpg  \n",
            "  inflating: data/cropped_faces/s45_02.jpg  \n",
            "  inflating: data/cropped_faces/s45_03.jpg  \n",
            "  inflating: data/cropped_faces/s45_04.jpg  \n",
            "  inflating: data/cropped_faces/s45_05.jpg  \n",
            "  inflating: data/cropped_faces/s45_06.jpg  \n",
            "  inflating: data/cropped_faces/s45_07.jpg  \n",
            "  inflating: data/cropped_faces/s45_08.jpg  \n",
            "  inflating: data/cropped_faces/s45_09.jpg  \n",
            "  inflating: data/cropped_faces/s45_10.jpg  \n",
            "  inflating: data/cropped_faces/s45_11.jpg  \n",
            "  inflating: data/cropped_faces/s45_12.jpg  \n",
            "  inflating: data/cropped_faces/s45_13.jpg  \n",
            "  inflating: data/cropped_faces/s45_14.jpg  \n",
            "  inflating: data/cropped_faces/s45_15.jpg  \n",
            "  inflating: data/cropped_faces/s46_01.jpg  \n",
            "  inflating: data/cropped_faces/s46_02.jpg  \n",
            "  inflating: data/cropped_faces/s46_03.jpg  \n",
            "  inflating: data/cropped_faces/s46_04.jpg  \n",
            "  inflating: data/cropped_faces/s46_05.jpg  \n",
            "  inflating: data/cropped_faces/s46_06.jpg  \n",
            "  inflating: data/cropped_faces/s46_07.jpg  \n",
            "  inflating: data/cropped_faces/s46_08.jpg  \n",
            "  inflating: data/cropped_faces/s46_09.jpg  \n",
            "  inflating: data/cropped_faces/s46_10.jpg  \n",
            "  inflating: data/cropped_faces/s46_11.jpg  \n",
            "  inflating: data/cropped_faces/s46_12.jpg  \n",
            "  inflating: data/cropped_faces/s46_13.jpg  \n",
            "  inflating: data/cropped_faces/s46_14.jpg  \n",
            "  inflating: data/cropped_faces/s46_15.jpg  \n",
            "  inflating: data/cropped_faces/s47_01.jpg  \n",
            "  inflating: data/cropped_faces/s47_02.jpg  \n",
            "  inflating: data/cropped_faces/s47_03.jpg  \n",
            "  inflating: data/cropped_faces/s47_04.jpg  \n",
            "  inflating: data/cropped_faces/s47_05.jpg  \n",
            "  inflating: data/cropped_faces/s47_06.jpg  \n",
            "  inflating: data/cropped_faces/s47_07.jpg  \n",
            "  inflating: data/cropped_faces/s47_08.jpg  \n",
            "  inflating: data/cropped_faces/s47_09.jpg  \n",
            "  inflating: data/cropped_faces/s47_10.jpg  \n",
            "  inflating: data/cropped_faces/s47_11.jpg  \n",
            "  inflating: data/cropped_faces/s47_12.jpg  \n",
            "  inflating: data/cropped_faces/s47_13.jpg  \n",
            "  inflating: data/cropped_faces/s47_14.jpg  \n",
            "  inflating: data/cropped_faces/s47_15.jpg  \n",
            "  inflating: data/cropped_faces/s48_01.jpg  \n",
            "  inflating: data/cropped_faces/s48_02.jpg  \n",
            "  inflating: data/cropped_faces/s48_03.jpg  \n",
            "  inflating: data/cropped_faces/s48_04.jpg  \n",
            "  inflating: data/cropped_faces/s48_05.jpg  \n",
            "  inflating: data/cropped_faces/s48_06.jpg  \n",
            "  inflating: data/cropped_faces/s48_07.jpg  \n",
            "  inflating: data/cropped_faces/s48_08.jpg  \n",
            "  inflating: data/cropped_faces/s48_09.jpg  \n",
            "  inflating: data/cropped_faces/s48_10.jpg  \n",
            "  inflating: data/cropped_faces/s48_11.jpg  \n",
            "  inflating: data/cropped_faces/s48_12.jpg  \n",
            "  inflating: data/cropped_faces/s48_13.jpg  \n",
            "  inflating: data/cropped_faces/s48_14.jpg  \n",
            "  inflating: data/cropped_faces/s48_15.jpg  \n",
            "  inflating: data/cropped_faces/s49_01.jpg  \n",
            "  inflating: data/cropped_faces/s49_02.jpg  \n",
            "  inflating: data/cropped_faces/s49_03.jpg  \n",
            "  inflating: data/cropped_faces/s49_04.jpg  \n",
            "  inflating: data/cropped_faces/s49_05.jpg  \n",
            "  inflating: data/cropped_faces/s49_06.jpg  \n",
            "  inflating: data/cropped_faces/s49_07.jpg  \n",
            "  inflating: data/cropped_faces/s49_08.jpg  \n",
            "  inflating: data/cropped_faces/s49_09.jpg  \n",
            "  inflating: data/cropped_faces/s49_10.jpg  \n",
            "  inflating: data/cropped_faces/s49_11.jpg  \n",
            "  inflating: data/cropped_faces/s49_12.jpg  \n",
            "  inflating: data/cropped_faces/s49_13.jpg  \n",
            "  inflating: data/cropped_faces/s49_14.jpg  \n",
            "  inflating: data/cropped_faces/s49_15.jpg  \n",
            "  inflating: data/cropped_faces/s50_01.jpg  \n",
            "  inflating: data/cropped_faces/s50_02.jpg  \n",
            "  inflating: data/cropped_faces/s50_03.jpg  \n",
            "  inflating: data/cropped_faces/s50_04.jpg  \n",
            "  inflating: data/cropped_faces/s50_05.jpg  \n",
            "  inflating: data/cropped_faces/s50_06.jpg  \n",
            "  inflating: data/cropped_faces/s50_07.jpg  \n",
            "  inflating: data/cropped_faces/s50_08.jpg  \n",
            "  inflating: data/cropped_faces/s50_09.jpg  \n",
            "  inflating: data/cropped_faces/s50_10.jpg  \n",
            "  inflating: data/cropped_faces/s50_11.jpg  \n",
            "  inflating: data/cropped_faces/s50_12.jpg  \n",
            "  inflating: data/cropped_faces/s50_13.jpg  \n",
            "  inflating: data/cropped_faces/s50_14.jpg  \n",
            "  inflating: data/cropped_faces/s50_15.jpg  \n"
          ]
        }
      ],
      "source": [
        "!unzip GTdb_crop.zip -d data # unzip the downloaded dataset"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import os\n",
        "import shutil\n",
        "import random\n",
        "import numpy as np\n",
        "import cv2\n",
        "from PIL import Image\n",
        "\n",
        "\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "\n",
        "import tensorflow as tf\n",
        "from tensorflow.keras import regularizers\n",
        "from tensorflow.keras.layers import Dense\n",
        "from tensorflow.keras.models import Model\n",
        "from tensorflow.keras.applications.vgg19 import VGG19\n",
        "\n",
        "from tensorflow.keras.applications.vgg19 import preprocess_input\n",
        "\n",
        "from tensorflow.keras.preprocessing.image import ImageDataGenerator\n",
        "from tensorflow.keras.models import Sequential\n",
        "from tensorflow.keras.layers import Conv2D, MaxPooling2D, Flatten, Dropout, GlobalAveragePooling2D\n",
        "\n",
        "from tensorflow.keras.utils import plot_model\n",
        "\n",
        "from tensorflow.keras.saving import save_model\n",
        "from tensorflow.keras.models import load_model\n",
        "\n",
        "from tensorflow.keras.preprocessing import image\n",
        "\n",
        "from sklearn.metrics.pairwise import cosine_similarity\n",
        "from sklearn.metrics import roc_curve, auc\n",
        "\n",
        "print(tf.version.VERSION)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Ej10YqM4uk6Q",
        "outputId": "26657b9e-f576-4b9a-ffe7-16366e11964e"
      },
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "2.12.0\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Define paths\n",
        "original_dir = '/content/data/cropped_faces'\n",
        "train_dir = '/content/process_data/train'\n",
        "val_dir = '/content/process_data/validation'\n",
        "test_dir = '/content/process_data/test'\n",
        "\n",
        "# Create train and validation folders\n",
        "os.makedirs(train_dir, exist_ok=True)\n",
        "os.makedirs(val_dir, exist_ok=True)\n",
        "os.makedirs(test_dir, exist_ok=True)\n",
        "\n",
        "# Define the number of images to put in each split\n",
        "train_count = 7\n",
        "val_count = 3\n",
        "test_count = 5\n",
        "\n",
        "# Loop through each person\n",
        "for i in range(1, 51):\n",
        "    person = \"s{:02d}\".format(i)\n",
        "    \n",
        "    # Create subdirectories for the person in each split\n",
        "    os.makedirs(os.path.join(train_dir, person), exist_ok=True)\n",
        "    os.makedirs(os.path.join(val_dir, person), exist_ok=True)\n",
        "    os.makedirs(os.path.join(test_dir, person), exist_ok=True)\n",
        "    \n",
        "    # Loop through each image for the person and copy it to the appropriate split\n",
        "    for j in range(1, 16):\n",
        "        image_name = \"{}_{:02d}.jpg\".format(person, j)\n",
        "        source_path = os.path.join(original_dir, image_name)\n",
        "        \n",
        "        if j <= train_count:\n",
        "            dest_path = os.path.join(train_dir, person, image_name)\n",
        "        elif j <= train_count + val_count:\n",
        "            dest_path = os.path.join(val_dir, person, image_name)\n",
        "        else:\n",
        "            dest_path = os.path.join(test_dir, person, image_name)\n",
        "        \n",
        "        shutil.copy(source_path, dest_path)"
      ],
      "metadata": {
        "id": "YfOlZrwcuxBd"
      },
      "execution_count": 3,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Define data generators\n",
        "train_datagen = ImageDataGenerator(\n",
        "    rescale=1./255,\n",
        "    rotation_range=20,  # rotate images by up to 20 degrees\n",
        "    width_shift_range=0.2,  # shift images horizontally by up to 20% of the image width\n",
        "    height_shift_range=0.2,  # shift images vertically by up to 20% of the image height\n",
        "    shear_range=0.2,  # apply shear transformation with a shear angle of up to 20 degrees\n",
        "    zoom_range=0.2,  # apply zoom transformation by up to 20%\n",
        "    horizontal_flip=True,  # flip images horizontally\n",
        "    fill_mode='nearest'  # fill any gaps created by the above transformations with the nearest pixel\n",
        ")\n",
        "val_datagen = ImageDataGenerator(\n",
        "    rescale=1./255,\n",
        "    rotation_range=20,  # rotate images by up to 20 degrees\n",
        "    width_shift_range=0.2,  # shift images horizontally by up to 20% of the image width\n",
        "    height_shift_range=0.2,  # shift images vertically by up to 20% of the image height\n",
        "    shear_range=0.2,  # apply shear transformation with a shear angle of up to 20 degrees\n",
        "    zoom_range=0.2,  # apply zoom transformation by up to 20%\n",
        "    horizontal_flip=True,  # flip images horizontally\n",
        "    fill_mode='nearest'  # fill any gaps created by the above transformations with the nearest pixel\n",
        ")\n",
        "\n",
        "\n",
        "train_generator = train_datagen.flow_from_directory(\n",
        "    train_dir,\n",
        "    target_size = (227,227),\n",
        "    batch_size=64,\n",
        "    class_mode='categorical')\n",
        "\n",
        "val_generator = val_datagen.flow_from_directory(\n",
        "    val_dir,\n",
        "    target_size = (227,227),\n",
        "    batch_size=64,\n",
        "    class_mode='categorical')"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "X_oLsIOTzIzw",
        "outputId": "505db919-8bfb-4e58-ee66-83e9b21ca683"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Found 350 images belonging to 50 classes.\n",
            "Found 150 images belonging to 50 classes.\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Load VGG19 model without top layers\n",
        "base_model = VGG19(weights='imagenet', include_top=False)\n",
        "\n",
        "# Add top layers\n",
        "x = base_model.output\n",
        "x1 = GlobalAveragePooling2D()(x)\n",
        "x2 = Dense(4096, activation='relu')(x1) # Keep the fully connected layer with 4096 units\n",
        "x3 = Dense(4096, activation='relu')(x2) # Keep the fully connected layer with 4096 units\n",
        "predictions = Dense(50, activation='softmax')(x3) # Change to 50 units for 50-class classifier\n",
        "\n",
        "# Create a new model with the added top layers\n",
        "model_1 = Model(inputs=base_model.input, outputs=predictions)\n",
        "\n",
        "# Freeze base model layers\n",
        "for layer in base_model.layers:\n",
        "    layer.trainable = False\n",
        "\n",
        "# Compile the model\n",
        "model_1.compile(optimizer='adam', loss='categorical_crossentropy', metrics=['accuracy'])\n",
        "\n",
        "# Print the model summary\n",
        "model_1.summary()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Ya10bd5pzqK5",
        "outputId": "6a402087-51ea-491b-f501-50d27d61196f"
      },
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Downloading data from https://storage.googleapis.com/tensorflow/keras-applications/vgg19/vgg19_weights_tf_dim_ordering_tf_kernels_notop.h5\n",
            "80134624/80134624 [==============================] - 1s 0us/step\n",
            "Model: \"model\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " input_1 (InputLayer)        [(None, None, None, 3)]   0         \n",
            "                                                                 \n",
            " block1_conv1 (Conv2D)       (None, None, None, 64)    1792      \n",
            "                                                                 \n",
            " block1_conv2 (Conv2D)       (None, None, None, 64)    36928     \n",
            "                                                                 \n",
            " block1_pool (MaxPooling2D)  (None, None, None, 64)    0         \n",
            "                                                                 \n",
            " block2_conv1 (Conv2D)       (None, None, None, 128)   73856     \n",
            "                                                                 \n",
            " block2_conv2 (Conv2D)       (None, None, None, 128)   147584    \n",
            "                                                                 \n",
            " block2_pool (MaxPooling2D)  (None, None, None, 128)   0         \n",
            "                                                                 \n",
            " block3_conv1 (Conv2D)       (None, None, None, 256)   295168    \n",
            "                                                                 \n",
            " block3_conv2 (Conv2D)       (None, None, None, 256)   590080    \n",
            "                                                                 \n",
            " block3_conv3 (Conv2D)       (None, None, None, 256)   590080    \n",
            "                                                                 \n",
            " block3_conv4 (Conv2D)       (None, None, None, 256)   590080    \n",
            "                                                                 \n",
            " block3_pool (MaxPooling2D)  (None, None, None, 256)   0         \n",
            "                                                                 \n",
            " block4_conv1 (Conv2D)       (None, None, None, 512)   1180160   \n",
            "                                                                 \n",
            " block4_conv2 (Conv2D)       (None, None, None, 512)   2359808   \n",
            "                                                                 \n",
            " block4_conv3 (Conv2D)       (None, None, None, 512)   2359808   \n",
            "                                                                 \n",
            " block4_conv4 (Conv2D)       (None, None, None, 512)   2359808   \n",
            "                                                                 \n",
            " block4_pool (MaxPooling2D)  (None, None, None, 512)   0         \n",
            "                                                                 \n",
            " block5_conv1 (Conv2D)       (None, None, None, 512)   2359808   \n",
            "                                                                 \n",
            " block5_conv2 (Conv2D)       (None, None, None, 512)   2359808   \n",
            "                                                                 \n",
            " block5_conv3 (Conv2D)       (None, None, None, 512)   2359808   \n",
            "                                                                 \n",
            " block5_conv4 (Conv2D)       (None, None, None, 512)   2359808   \n",
            "                                                                 \n",
            " block5_pool (MaxPooling2D)  (None, None, None, 512)   0         \n",
            "                                                                 \n",
            " global_average_pooling2d (G  (None, 512)              0         \n",
            " lobalAveragePooling2D)                                          \n",
            "                                                                 \n",
            " dense (Dense)               (None, 4096)              2101248   \n",
            "                                                                 \n",
            " dense_1 (Dense)             (None, 4096)              16781312  \n",
            "                                                                 \n",
            " dense_2 (Dense)             (None, 50)                204850    \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 39,111,794\n",
            "Trainable params: 19,087,410\n",
            "Non-trainable params: 20,024,384\n",
            "_________________________________________________________________\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Train the model\n",
        "history_1 = model_1.fit(\n",
        "    train_generator,\n",
        "    epochs=20,  # specify the number of epochs for training\n",
        "    validation_data=val_generator\n",
        ")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "dEX78O4h2eNv",
        "outputId": "862b2517-9b8e-4b61-83bd-29cbb42e9ba0"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/20\n",
            "6/6 [==============================] - 412s 72s/step - loss: 4.5649 - accuracy: 0.0171 - val_loss: 3.9379 - val_accuracy: 0.0400\n",
            "Epoch 2/20\n",
            "6/6 [==============================] - 412s 72s/step - loss: 3.9611 - accuracy: 0.0229 - val_loss: 3.9044 - val_accuracy: 0.0533\n",
            "Epoch 3/20\n",
            "6/6 [==============================] - 419s 79s/step - loss: 3.9117 - accuracy: 0.0171 - val_loss: 3.8924 - val_accuracy: 0.0333\n",
            "Epoch 4/20\n",
            "6/6 [==============================] - 411s 72s/step - loss: 3.8866 - accuracy: 0.0371 - val_loss: 3.8609 - val_accuracy: 0.0333\n",
            "Epoch 5/20\n",
            "6/6 [==============================] - 411s 72s/step - loss: 3.8420 - accuracy: 0.0743 - val_loss: 3.8042 - val_accuracy: 0.0733\n",
            "Epoch 6/20\n",
            "6/6 [==============================] - 416s 72s/step - loss: 3.7698 - accuracy: 0.0600 - val_loss: 3.7550 - val_accuracy: 0.0533\n",
            "Epoch 7/20\n",
            "6/6 [==============================] - 416s 73s/step - loss: 3.6302 - accuracy: 0.0971 - val_loss: 3.5542 - val_accuracy: 0.1067\n",
            "Epoch 8/20\n",
            "6/6 [==============================] - 411s 72s/step - loss: 3.4899 - accuracy: 0.0914 - val_loss: 3.4531 - val_accuracy: 0.1000\n",
            "Epoch 9/20\n",
            "6/6 [==============================] - 413s 72s/step - loss: 3.4178 - accuracy: 0.1143 - val_loss: 3.4015 - val_accuracy: 0.1000\n",
            "Epoch 10/20\n",
            "6/6 [==============================] - 416s 78s/step - loss: 3.3561 - accuracy: 0.0857 - val_loss: 3.3302 - val_accuracy: 0.1067\n",
            "Epoch 11/20\n",
            "6/6 [==============================] - 410s 72s/step - loss: 3.2529 - accuracy: 0.1200 - val_loss: 3.3707 - val_accuracy: 0.1400\n",
            "Epoch 12/20\n",
            "6/6 [==============================] - 412s 72s/step - loss: 3.1427 - accuracy: 0.1314 - val_loss: 3.4645 - val_accuracy: 0.0867\n",
            "Epoch 13/20\n",
            "6/6 [==============================] - 416s 79s/step - loss: 3.0475 - accuracy: 0.1857 - val_loss: 3.2062 - val_accuracy: 0.1133\n",
            "Epoch 14/20\n",
            "6/6 [==============================] - 411s 72s/step - loss: 3.0244 - accuracy: 0.1657 - val_loss: 3.0752 - val_accuracy: 0.1533\n",
            "Epoch 15/20\n",
            "6/6 [==============================] - 411s 77s/step - loss: 2.9065 - accuracy: 0.1943 - val_loss: 3.0741 - val_accuracy: 0.1800\n",
            "Epoch 16/20\n",
            "6/6 [==============================] - 417s 72s/step - loss: 2.7446 - accuracy: 0.2429 - val_loss: 3.0084 - val_accuracy: 0.1667\n",
            "Epoch 17/20\n",
            "6/6 [==============================] - 390s 68s/step - loss: 2.6611 - accuracy: 0.2343 - val_loss: 3.1382 - val_accuracy: 0.1733\n",
            "Epoch 18/20\n",
            "6/6 [==============================] - 414s 73s/step - loss: 2.6662 - accuracy: 0.2257 - val_loss: 2.8904 - val_accuracy: 0.2133\n",
            "Epoch 19/20\n",
            "6/6 [==============================] - 400s 75s/step - loss: 2.4748 - accuracy: 0.3029 - val_loss: 2.8277 - val_accuracy: 0.2133\n",
            "Epoch 20/20\n",
            "6/6 [==============================] - 407s 71s/step - loss: 2.4359 - accuracy: 0.3229 - val_loss: 2.9290 - val_accuracy: 0.2667\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Plot the training and validation accuracy\n",
        "plt.plot(history_1.history['accuracy'])\n",
        "plt.plot(history_1.history['val_accuracy'])\n",
        "plt.title('Model Accuracy')\n",
        "plt.ylabel('Accuracy')\n",
        "plt.xlabel('Epoch')\n",
        "plt.legend(['train', 'val'], loc='upper left')\n",
        "plt.show()\n",
        "\n",
        "# Plot the training and validation loss\n",
        "plt.plot(history_1.history['loss'])\n",
        "plt.plot(history_1.history['val_loss'])\n",
        "plt.title('Model Loss')\n",
        "plt.ylabel('Loss')\n",
        "plt.xlabel('Epoch')\n",
        "plt.legend(['train', 'val'], loc='upper left')\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 927
        },
        "id": "NAqsxRZJ2szq",
        "outputId": "ff4bcb24-7855-4ab6-a52c-80decd0ce4dc"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "model_1.save_weights('model_1_weights.h5')"
      ],
      "metadata": {
        "id": "WO9PhzQ15wSA"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "model_1.save('model_1.h5')"
      ],
      "metadata": {
        "id": "5yM5TdlpXEBO"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Load the model from the H5 file\n",
        "model_1.load_weights('/content/model_1_weights.h5')"
      ],
      "metadata": {
        "id": "EZRYTTdBpkYi"
      },
      "execution_count": 5,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Create a new model that outputs the fc7 features\n",
        "fc7_layer = model_1.get_layer('dense_1')\n",
        "fc7_output = fc7_layer.output\n",
        "model_1_withoudHead = Model(inputs=model_1.input, outputs=fc7_output)"
      ],
      "metadata": {
        "id": "DBaIegZ6YeZI"
      },
      "execution_count": 6,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Define the paths to the validation dataset\n",
        "test_dir = '/content/process_data/test'\n",
        "\n",
        "# Define the batch size for validation data\n",
        "batch_size = 1\n",
        "\n",
        "# Create an empty list to hold the feature vectors for the validation images\n",
        "test_features = []\n",
        "\n",
        "# Loop through the validation image directories\n",
        "for folder_name in os.listdir(test_dir):\n",
        "\n",
        "    test_directory = os.path.join(test_dir, folder_name)\n",
        "\n",
        "    # Create an empty list to hold the feature vectors for the validation images in the current directory\n",
        "    directory_features = []\n",
        "    \n",
        "    # Loop through the validation images in the current directory\n",
        "    for i in range(11, 16):\n",
        "        # Load the validation image using PIL\n",
        "        img_path = os.path.join(test_directory, f'{folder_name}_{i}.jpg')\n",
        "        img = image.load_img(img_path, target_size=(227, 227))\n",
        "        \n",
        "        # Convert the image to a numpy array and preprocess it\n",
        "        x = image.img_to_array(img)\n",
        "        x = np.expand_dims(x, axis=0)\n",
        "        x = preprocess_input(x)\n",
        "        \n",
        "        # Extract the features using the VGG19 model\n",
        "        features = model_1_withoudHead.predict(x)\n",
        "        \n",
        "        # Add the features to the list of features for the current directory\n",
        "        directory_features.append(features.flatten())\n",
        "    \n",
        "    # Convert the list of features for the current directory to a numpy array\n",
        "    directory_features = np.array(directory_features)\n",
        "    \n",
        "    # Add the numpy array to the list of feature vectors for all validation images\n",
        "    test_features.append(directory_features)\n",
        "\n",
        "# Convert the list of feature vectors for all validation images to a numpy array\n",
        "test_features = np.array(test_features)\n",
        "print(test_features)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Xc6FmLqdgUsz",
        "outputId": "a55a30e7-ee30-4406-daea-51ac012f664e"
      },
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 655ms/step\n",
            "1/1 [==============================] - 1s 644ms/step\n",
            "1/1 [==============================] - 1s 650ms/step\n",
            "1/1 [==============================] - 1s 614ms/step\n",
            "1/1 [==============================] - 1s 635ms/step\n",
            "1/1 [==============================] - 1s 671ms/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 869ms/step\n",
            "1/1 [==============================] - 1s 624ms/step\n",
            "1/1 [==============================] - 1s 632ms/step\n",
            "1/1 [==============================] - 1s 632ms/step\n",
            "1/1 [==============================] - 1s 633ms/step\n",
            "1/1 [==============================] - 1s 655ms/step\n",
            "1/1 [==============================] - 1s 646ms/step\n",
            "1/1 [==============================] - 1s 624ms/step\n",
            "1/1 [==============================] - 1s 628ms/step\n",
            "1/1 [==============================] - 1s 641ms/step\n",
            "1/1 [==============================] - 1s 619ms/step\n",
            "1/1 [==============================] - 1s 620ms/step\n",
            "1/1 [==============================] - 1s 629ms/step\n",
            "1/1 [==============================] - 1s 632ms/step\n",
            "1/1 [==============================] - 1s 616ms/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 930ms/step\n",
            "1/1 [==============================] - 1s 635ms/step\n",
            "1/1 [==============================] - 1s 635ms/step\n",
            "1/1 [==============================] - 1s 626ms/step\n",
            "1/1 [==============================] - 1s 634ms/step\n",
            "1/1 [==============================] - 1s 628ms/step\n",
            "1/1 [==============================] - 1s 630ms/step\n",
            "1/1 [==============================] - 1s 638ms/step\n",
            "1/1 [==============================] - 1s 636ms/step\n",
            "1/1 [==============================] - 1s 628ms/step\n",
            "1/1 [==============================] - 1s 631ms/step\n",
            "1/1 [==============================] - 1s 637ms/step\n",
            "1/1 [==============================] - 1s 617ms/step\n",
            "1/1 [==============================] - 1s 639ms/step\n",
            "1/1 [==============================] - 1s 633ms/step\n",
            "1/1 [==============================] - 1s 893ms/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 631ms/step\n",
            "1/1 [==============================] - 1s 626ms/step\n",
            "1/1 [==============================] - 1s 628ms/step\n",
            "1/1 [==============================] - 1s 630ms/step\n",
            "1/1 [==============================] - 1s 650ms/step\n",
            "1/1 [==============================] - 1s 627ms/step\n",
            "1/1 [==============================] - 1s 667ms/step\n",
            "1/1 [==============================] - 1s 655ms/step\n",
            "1/1 [==============================] - 1s 658ms/step\n",
            "1/1 [==============================] - 1s 662ms/step\n",
            "1/1 [==============================] - 1s 648ms/step\n",
            "1/1 [==============================] - 1s 664ms/step\n",
            "1/1 [==============================] - 1s 657ms/step\n",
            "1/1 [==============================] - 1s 649ms/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 658ms/step\n",
            "1/1 [==============================] - 1s 654ms/step\n",
            "1/1 [==============================] - 1s 667ms/step\n",
            "1/1 [==============================] - 1s 650ms/step\n",
            "1/1 [==============================] - 1s 663ms/step\n",
            "1/1 [==============================] - 1s 673ms/step\n",
            "1/1 [==============================] - 1s 669ms/step\n",
            "1/1 [==============================] - 1s 658ms/step\n",
            "1/1 [==============================] - 1s 684ms/step\n",
            "1/1 [==============================] - 1s 669ms/step\n",
            "1/1 [==============================] - 1s 663ms/step\n",
            "1/1 [==============================] - 1s 670ms/step\n",
            "1/1 [==============================] - 1s 662ms/step\n",
            "1/1 [==============================] - 1s 716ms/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 911ms/step\n",
            "1/1 [==============================] - 1s 654ms/step\n",
            "1/1 [==============================] - 1s 686ms/step\n",
            "1/1 [==============================] - 1s 676ms/step\n",
            "1/1 [==============================] - 1s 675ms/step\n",
            "1/1 [==============================] - 1s 661ms/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 969ms/step\n",
            "1/1 [==============================] - 1s 681ms/step\n",
            "1/1 [==============================] - 1s 679ms/step\n",
            "1/1 [==============================] - 1s 666ms/step\n",
            "1/1 [==============================] - 1s 660ms/step\n",
            "1/1 [==============================] - 1s 676ms/step\n",
            "1/1 [==============================] - 1s 928ms/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 692ms/step\n",
            "1/1 [==============================] - 1s 665ms/step\n",
            "1/1 [==============================] - 1s 653ms/step\n",
            "1/1 [==============================] - 1s 659ms/step\n",
            "1/1 [==============================] - 1s 666ms/step\n",
            "1/1 [==============================] - 1s 656ms/step\n",
            "1/1 [==============================] - 1s 656ms/step\n",
            "1/1 [==============================] - 1s 672ms/step\n",
            "1/1 [==============================] - 1s 671ms/step\n",
            "1/1 [==============================] - 1s 663ms/step\n",
            "1/1 [==============================] - 1s 684ms/step\n",
            "1/1 [==============================] - 1s 681ms/step\n",
            "1/1 [==============================] - 1s 656ms/step\n",
            "1/1 [==============================] - 1s 670ms/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 954ms/step\n",
            "1/1 [==============================] - 1s 679ms/step\n",
            "1/1 [==============================] - 1s 655ms/step\n",
            "1/1 [==============================] - 1s 665ms/step\n",
            "1/1 [==============================] - 1s 651ms/step\n",
            "1/1 [==============================] - 1s 667ms/step\n",
            "1/1 [==============================] - 1s 693ms/step\n",
            "1/1 [==============================] - 1s 659ms/step\n",
            "1/1 [==============================] - 1s 666ms/step\n",
            "1/1 [==============================] - 1s 671ms/step\n",
            "1/1 [==============================] - 1s 671ms/step\n",
            "1/1 [==============================] - 1s 677ms/step\n",
            "1/1 [==============================] - 1s 663ms/step\n",
            "1/1 [==============================] - 1s 683ms/step\n",
            "1/1 [==============================] - 1s 912ms/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 702ms/step\n",
            "1/1 [==============================] - 1s 656ms/step\n",
            "1/1 [==============================] - 1s 682ms/step\n",
            "1/1 [==============================] - 1s 659ms/step\n",
            "1/1 [==============================] - 1s 661ms/step\n",
            "1/1 [==============================] - 1s 668ms/step\n",
            "1/1 [==============================] - 1s 671ms/step\n",
            "1/1 [==============================] - 1s 651ms/step\n",
            "1/1 [==============================] - 1s 670ms/step\n",
            "1/1 [==============================] - 1s 698ms/step\n",
            "1/1 [==============================] - 1s 659ms/step\n",
            "1/1 [==============================] - 1s 667ms/step\n",
            "1/1 [==============================] - 1s 665ms/step\n",
            "1/1 [==============================] - 1s 667ms/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 666ms/step\n",
            "1/1 [==============================] - 1s 650ms/step\n",
            "1/1 [==============================] - 1s 664ms/step\n",
            "1/1 [==============================] - 1s 687ms/step\n",
            "1/1 [==============================] - 1s 668ms/step\n",
            "1/1 [==============================] - 1s 656ms/step\n",
            "1/1 [==============================] - 1s 680ms/step\n",
            "1/1 [==============================] - 1s 665ms/step\n",
            "1/1 [==============================] - 1s 655ms/step\n",
            "1/1 [==============================] - 1s 668ms/step\n",
            "1/1 [==============================] - 1s 695ms/step\n",
            "1/1 [==============================] - 1s 686ms/step\n",
            "1/1 [==============================] - 1s 654ms/step\n",
            "1/1 [==============================] - 1s 974ms/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 656ms/step\n",
            "1/1 [==============================] - 1s 655ms/step\n",
            "1/1 [==============================] - 1s 678ms/step\n",
            "1/1 [==============================] - 1s 649ms/step\n",
            "1/1 [==============================] - 1s 668ms/step\n",
            "1/1 [==============================] - 1s 664ms/step\n",
            "1/1 [==============================] - 1s 661ms/step\n",
            "1/1 [==============================] - 1s 649ms/step\n",
            "1/1 [==============================] - 1s 658ms/step\n",
            "1/1 [==============================] - 1s 664ms/step\n",
            "1/1 [==============================] - 1s 655ms/step\n",
            "1/1 [==============================] - 1s 636ms/step\n",
            "1/1 [==============================] - 1s 637ms/step\n",
            "1/1 [==============================] - 1s 644ms/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 805ms/step\n",
            "1/1 [==============================] - 1s 628ms/step\n",
            "1/1 [==============================] - 1s 641ms/step\n",
            "1/1 [==============================] - 1s 624ms/step\n",
            "1/1 [==============================] - 1s 625ms/step\n",
            "1/1 [==============================] - 1s 631ms/step\n",
            "1/1 [==============================] - 1s 620ms/step\n",
            "1/1 [==============================] - 1s 628ms/step\n",
            "1/1 [==============================] - 1s 643ms/step\n",
            "1/1 [==============================] - 1s 628ms/step\n",
            "1/1 [==============================] - 1s 624ms/step\n",
            "1/1 [==============================] - 1s 633ms/step\n",
            "1/1 [==============================] - 1s 644ms/step\n",
            "1/1 [==============================] - 1s 652ms/step\n",
            "1/1 [==============================] - 1s 623ms/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 913ms/step\n",
            "1/1 [==============================] - 1s 624ms/step\n",
            "1/1 [==============================] - 1s 633ms/step\n",
            "1/1 [==============================] - 1s 638ms/step\n",
            "1/1 [==============================] - 1s 621ms/step\n",
            "1/1 [==============================] - 1s 618ms/step\n",
            "1/1 [==============================] - 1s 636ms/step\n",
            "1/1 [==============================] - 1s 627ms/step\n",
            "1/1 [==============================] - 1s 645ms/step\n",
            "1/1 [==============================] - 1s 652ms/step\n",
            "1/1 [==============================] - 1s 633ms/step\n",
            "1/1 [==============================] - 1s 620ms/step\n",
            "1/1 [==============================] - 1s 629ms/step\n",
            "1/1 [==============================] - 1s 642ms/step\n",
            "1/1 [==============================] - 1s 634ms/step\n",
            "1/1 [==============================] - 1s 981ms/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 954ms/step\n",
            "1/1 [==============================] - 1s 635ms/step\n",
            "1/1 [==============================] - 1s 642ms/step\n",
            "1/1 [==============================] - 1s 621ms/step\n",
            "1/1 [==============================] - 1s 627ms/step\n",
            "1/1 [==============================] - 1s 649ms/step\n",
            "1/1 [==============================] - 1s 612ms/step\n",
            "1/1 [==============================] - 1s 629ms/step\n",
            "1/1 [==============================] - 1s 629ms/step\n",
            "1/1 [==============================] - 1s 628ms/step\n",
            "1/1 [==============================] - 1s 647ms/step\n",
            "1/1 [==============================] - 1s 628ms/step\n",
            "1/1 [==============================] - 1s 638ms/step\n",
            "1/1 [==============================] - 1s 628ms/step\n",
            "1/1 [==============================] - 1s 641ms/step\n",
            "1/1 [==============================] - 1s 948ms/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 988ms/step\n",
            "1/1 [==============================] - 1s 663ms/step\n",
            "1/1 [==============================] - 1s 637ms/step\n",
            "1/1 [==============================] - 1s 635ms/step\n",
            "1/1 [==============================] - 1s 676ms/step\n",
            "1/1 [==============================] - 1s 628ms/step\n",
            "1/1 [==============================] - 1s 644ms/step\n",
            "1/1 [==============================] - 1s 629ms/step\n",
            "1/1 [==============================] - 1s 628ms/step\n",
            "1/1 [==============================] - 1s 627ms/step\n",
            "1/1 [==============================] - 1s 628ms/step\n",
            "1/1 [==============================] - 1s 625ms/step\n",
            "1/1 [==============================] - 1s 636ms/step\n",
            "1/1 [==============================] - 1s 649ms/step\n",
            "1/1 [==============================] - 1s 616ms/step\n",
            "1/1 [==============================] - 1s 911ms/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 619ms/step\n",
            "1/1 [==============================] - 1s 626ms/step\n",
            "1/1 [==============================] - 1s 627ms/step\n",
            "1/1 [==============================] - 1s 626ms/step\n",
            "1/1 [==============================] - 1s 646ms/step\n",
            "1/1 [==============================] - 1s 621ms/step\n",
            "[[[ 0.        14.139654   2.4502354 ...  0.         0.         0.       ]\n",
            "  [ 0.        12.290299   2.2040577 ...  0.         0.         0.       ]\n",
            "  [ 0.        11.337109   3.6494446 ...  0.         0.         0.       ]\n",
            "  [ 0.         9.529051   1.8174248 ...  0.         0.         0.       ]\n",
            "  [ 0.        12.327253   2.29327   ...  0.         0.         0.       ]]\n",
            "\n",
            " [[ 0.        12.937393   2.1804833 ...  0.         0.         0.       ]\n",
            "  [ 0.        17.155766   4.6214356 ...  0.         0.         0.       ]\n",
            "  [ 0.        15.967076   5.600358  ...  0.         0.         0.       ]\n",
            "  [ 0.        15.440191   5.507016  ...  0.         0.         0.       ]\n",
            "  [ 0.        13.943959   3.382998  ...  0.         0.         0.       ]]\n",
            "\n",
            " [[ 0.        10.918902   2.3112633 ...  0.         0.         0.       ]\n",
            "  [ 0.        12.825713   4.763537  ...  0.         0.         0.       ]\n",
            "  [ 0.        11.063464   3.4241288 ...  0.         0.         0.       ]\n",
            "  [ 0.        11.023556   3.23294   ...  0.         0.         0.       ]\n",
            "  [ 0.        11.564325   4.233281  ...  0.         0.         0.       ]]\n",
            "\n",
            " ...\n",
            "\n",
            " [[ 0.         8.852306   1.7801068 ...  0.         0.         0.       ]\n",
            "  [ 0.        11.645975   1.905397  ...  0.         0.         0.       ]\n",
            "  [ 0.         9.69101    2.652676  ...  0.         0.         0.       ]\n",
            "  [ 0.         9.305297   1.922468  ...  0.         0.         0.       ]\n",
            "  [ 0.        10.307397   1.7868052 ...  0.         0.         0.       ]]\n",
            "\n",
            " [[ 0.        20.513731   5.804276  ...  0.         0.         0.       ]\n",
            "  [ 0.        18.071182   3.333413  ...  0.         0.         0.       ]\n",
            "  [ 0.         9.993474   5.3166237 ...  0.         0.         0.       ]\n",
            "  [ 0.        12.485815   6.2111435 ...  0.         0.         0.       ]\n",
            "  [ 0.        11.403346   3.7059228 ...  0.         0.         0.       ]]\n",
            "\n",
            " [[ 0.        11.138346   2.2877529 ...  0.         0.         0.       ]\n",
            "  [ 0.        11.691745   2.56546   ...  0.         0.         0.       ]\n",
            "  [ 0.        11.99364    2.314498  ...  0.         0.         0.       ]\n",
            "  [ 0.        13.131907   2.240363  ...  0.         0.         0.       ]\n",
            "  [ 0.        14.397628   2.1494749 ...  0.         0.         0.       ]]]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "print(test_features.shape)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "djZsHkgF7qDV",
        "outputId": "8d5df7ff-bff5-497f-bd66-9c9a20790495"
      },
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(50, 5, 4096)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Create a list for enrollment features for all subjects\n",
        "enrollment_features = [[0] * 4096 for i in range(50)]\n",
        "\n",
        "# Create a list for verification features for all subjects and each verification image\n",
        "verification_features = [[[0] * 4096 for k in range(4)] for j in range(50)]\n",
        "\n",
        "# Populate the enrollment_features list\n",
        "for i in range(0,50):\n",
        "    enrollment_features[i] = test_features[i][0]\n",
        "\n",
        "# Populate the verification_features list\n",
        "for j in range(0,50):\n",
        "    verification_features[j][0] = test_features[j][1]\n",
        "    verification_features[j][1] = test_features[j][2]\n",
        "    verification_features[j][2] = test_features[j][3]\n",
        "    verification_features[j][3] = test_features[j][4]\n",
        "\n",
        "# Create a numpy array to store the cosine similarities\n",
        "cosine_similarities = np.zeros((10000, 2))\n",
        "\n",
        "# Iterate over each person and their enrollment and verification images\n",
        "for i in range(0,50):\n",
        "    a = 200 * i\n",
        "    for j in range(0,50):\n",
        "        b = a + (4 * j)\n",
        "        \n",
        "        # Get the enrollment feature and the four verification features\n",
        "        enrollment_features_2d = enrollment_features[i].reshape(1, -1)\n",
        "        verification_features_1_2d = verification_features[j][0].reshape(1, -1)\n",
        "        verification_features_2_2d = verification_features[j][1].reshape(1, -1)\n",
        "        verification_features_3_2d = verification_features[j][2].reshape(1, -1)\n",
        "        verification_features_4_2d = verification_features[j][3].reshape(1, -1)\n",
        "        # Calculate the cosine similarity between the enrollment and verification features\n",
        "        cosine_similarities[b][0] = cosine_similarity(enrollment_features_2d, verification_features_1_2d)\n",
        "        cosine_similarities[b+1][0] = cosine_similarity(enrollment_features_2d, verification_features_2_2d)\n",
        "        cosine_similarities[b+2][0] = cosine_similarity(enrollment_features_2d, verification_features_3_2d)\n",
        "        cosine_similarities[b+3][0] = cosine_similarity(enrollment_features_2d, verification_features_4_2d)\n",
        "        # Set the cosine similarity to 1 if the enrollment and verification images are from the same person\n",
        "        if i == j:\n",
        "            cosine_similarities[b][1] = 1\n",
        "            cosine_similarities[b+1][1] = 1\n",
        "            cosine_similarities[b+2][1] = 1\n",
        "            cosine_similarities[b+3][1] = 1\n",
        "\n",
        "print(cosine_similarities)\n",
        "print(cosine_similarities.shape)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "zoCAEpAuiEif",
        "outputId": "431ead6c-183c-4b0a-8405-a7c81f56f5a2"
      },
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[[0.99692619 1.        ]\n",
            " [0.9806056  1.        ]\n",
            " [0.99615943 1.        ]\n",
            " ...\n",
            " [0.99809468 1.        ]\n",
            " [0.9980765  1.        ]\n",
            " [0.99494362 1.        ]]\n",
            "(10000, 2)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Transpose and extract all cosine similarities\n",
        "cosine_similarities_t = np.transpose(cosine_similarities)\n",
        "cosine_similarities_raw = cosine_similarities_t[0]\n",
        "\n",
        "# Plot the histogram\n",
        "plt.hist(cosine_similarities_raw, bins=50)\n",
        "plt.xlabel('Cosine similarity')\n",
        "plt.ylabel('Count')\n",
        "plt.title('Histogram of cosine similarities')\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 472
        },
        "id": "QHoBDOU9mel_",
        "outputId": "99b6377f-edf5-4cf6-9518-27a77fe7f0b5"
      },
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Create emply list for genuine and imposter scores\n",
        "genuine_scores = []\n",
        "imposter_scores = []\n",
        "\n",
        "# Populate the lists based on true labels provided\n",
        "for i in range(cosine_similarities.shape[0]):\n",
        "    if cosine_similarities[i][1] == 1:\n",
        "        genuine_scores.append(cosine_similarities[i][0])\n",
        "    else:\n",
        "        imposter_scores.append(cosine_similarities[i][0])"
      ],
      "metadata": {
        "id": "kklIw2gfmiAJ"
      },
      "execution_count": 11,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Plot histogram for genuine_scores\n",
        "plt.hist(genuine_scores, bins=50, label='Genuine Scores')\n",
        "plt.title('Histogram of Genuine Scores')\n",
        "plt.xlabel('Cosine Similarity')\n",
        "plt.ylabel('Frequency')\n",
        "plt.legend(loc='upper right')\n",
        "plt.show()\n",
        "\n",
        "# Plot histogram for imposter_scores\n",
        "plt.hist(imposter_scores, bins=50, label='Imposter Scores')\n",
        "plt.title('Histogram of Imposter Scores')\n",
        "plt.xlabel('Cosine Similarity')\n",
        "plt.ylabel('Frequency')\n",
        "plt.legend(loc='upper right')\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 927
        },
        "id": "XM96xnurnA16",
        "outputId": "d6321af4-e806-41e1-d0b4-eae4666af4d0"
      },
      "execution_count": 12,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# create arrays to store true labels and similarity scores\n",
        "true_labels = cosine_similarities_t[1]\n",
        "scores = cosine_similarities_t[0]\n",
        "\n",
        "# create arrays to store FPR and TPR for different thresholds\n",
        "fpr = []\n",
        "tpr = []\n",
        "thresholds = []\n",
        "\n",
        "# calculate the FPR and TPR for different thresholds\n",
        "fpr, tpr, thresholds = roc_curve(true_labels, scores)\n",
        "\n",
        "# calculate the area under the ROC curve (AUC)\n",
        "auc_score = auc(fpr, tpr)\n",
        "\n",
        "# calculate d prime\n",
        "d_prime = np.abs(np.mean(scores[true_labels == 1]) - np.mean(scores[true_labels == 0])) / np.sqrt(0.5 * (np.std(scores[true_labels == 1]) ** 2 + np.std(scores[true_labels == 0]) ** 2))\n",
        "\n",
        "# Calculate the optimal operating point on the ROC curve\n",
        "optimal_idx = np.argmax(tpr - fpr)\n",
        "OPTROCPT = (fpr[optimal_idx], tpr[optimal_idx])\n",
        "\n",
        "# plot the ROC curve\n",
        "plt.plot(fpr, tpr, color='darkorange', lw=2, label='ROC curve (AUC = %0.4f, d\\' = %0.4f)' % (auc_score, d_prime))\n",
        "plt.plot(OPTROCPT[0], OPTROCPT[1], 'ro', label='Optimal Operating Point (fpr = %0.4f, tpr = %0.4f)' % (fpr[optimal_idx], tpr[optimal_idx]))\n",
        "plt.plot([0, 1], [0, 1], color='navy', lw=2, linestyle='--')\n",
        "plt.xlim([0.0, 1.0])\n",
        "plt.ylim([0.0, 1.0])\n",
        "plt.xlabel('False Positive Rate')\n",
        "plt.ylabel('True Positive Rate')\n",
        "plt.title('Receiver operating characteristic')\n",
        "plt.legend(loc=\"lower right\")\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 472
        },
        "id": "bPHL2IlksZi_",
        "outputId": "7f329c95-34e7-46ce-9cdc-2ae5cf11b39f"
      },
      "execution_count": 13,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "original_data = cosine_similarities\n",
        "\n",
        "# Define parameters\n",
        "gallery_size = 200  # Number of rows to include for each probe\n",
        "probe_size = 50  # Total number of probes (rows in the new object)\n",
        "\n",
        "# Initialize the new object for the rank1 and rank5 identification calculations\n",
        "cosine_similarities_probes = np.empty((probe_size, gallery_size, 2))\n",
        "\n",
        "# Loop through the chunks\n",
        "for i in range(probe_size):\n",
        "    # Extract the corresponding chunk of rows from the original dataset\n",
        "    start_index = i * gallery_size\n",
        "    end_index = start_index + gallery_size\n",
        "    chunk = original_data[start_index:end_index, :]\n",
        "\n",
        "    # Sort the chunk based on the first column in descending order\n",
        "    sorted_chunk = chunk[chunk[:, 0].argsort()[::-1]]\n",
        "\n",
        "    # Store the sorted chunk in the new object\n",
        "    cosine_similarities_probes[i] = sorted_chunk\n",
        "\n",
        "# Print the new object\n",
        "print(cosine_similarities_probes)\n",
        "print(cosine_similarities_probes.shape)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "yu3ycyUJbn4t",
        "outputId": "a498a192-102a-4f52-8173-cf860c88ba82"
      },
      "execution_count": 14,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[[[0.99873084 0.        ]\n",
            "  [0.99843401 0.        ]\n",
            "  [0.99839449 0.        ]\n",
            "  ...\n",
            "  [0.79676086 0.        ]\n",
            "  [0.70563    0.        ]\n",
            "  [0.67561495 0.        ]]\n",
            "\n",
            " [[0.99852604 0.        ]\n",
            "  [0.99831593 0.        ]\n",
            "  [0.9980408  0.        ]\n",
            "  ...\n",
            "  [0.80375898 0.        ]\n",
            "  [0.71280944 0.        ]\n",
            "  [0.68357146 0.        ]]\n",
            "\n",
            " [[0.99796569 0.        ]\n",
            "  [0.99764335 0.        ]\n",
            "  [0.99756992 0.        ]\n",
            "  ...\n",
            "  [0.81068146 0.        ]\n",
            "  [0.72273338 0.        ]\n",
            "  [0.69157279 0.        ]]\n",
            "\n",
            " ...\n",
            "\n",
            " [[0.9981389  1.        ]\n",
            "  [0.99794191 0.        ]\n",
            "  [0.99770689 0.        ]\n",
            "  ...\n",
            "  [0.84931546 0.        ]\n",
            "  [0.76498139 0.        ]\n",
            "  [0.74002254 0.        ]]\n",
            "\n",
            " [[0.9971686  0.        ]\n",
            "  [0.99678552 0.        ]\n",
            "  [0.99639165 0.        ]\n",
            "  ...\n",
            "  [0.87040675 0.        ]\n",
            "  [0.82879907 0.        ]\n",
            "  [0.79481399 0.        ]]\n",
            "\n",
            " [[0.99861813 0.        ]\n",
            "  [0.99830443 0.        ]\n",
            "  [0.9983021  0.        ]\n",
            "  ...\n",
            "  [0.84593242 0.        ]\n",
            "  [0.75999165 0.        ]\n",
            "  [0.73442757 0.        ]]]\n",
            "(50, 200, 2)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Initialize the new object for the rank1 and rank5 identifications\n",
        "rank1_identification = np.zeros(50)\n",
        "rank5_identification = np.zeros(50)\n",
        "\n",
        "# Initialize the new object for the rank1 and rank5 identifications using the optimal threshhold\n",
        "rank1_identification_withThreshold = np.zeros(50)\n",
        "rank5_identification_withThreshold = np.zeros(50)\n",
        "\n",
        "# Iterate through the first axis of cosine_similarities_probes\n",
        "for i in range(50):\n",
        "    # Select the first element and retrieve the label of that selection\n",
        "    rank1_identification[i] = (cosine_similarities_probes[i][0][1])\n",
        "\n",
        "    # Iterate through the second axis of cosine_similarities_probes\n",
        "    for j in range(5):\n",
        "        # Retrieve the positive label if there is one\n",
        "        if cosine_similarities_probes[i][j][1] == 1:\n",
        "            rank5_identification[i] = (cosine_similarities_probes[i][j][1])\n",
        "\n",
        "# Setting the threshhold for new rank1 and rank5 identifications\n",
        "optimal_threshhold = fpr[optimal_idx]\n",
        "\n",
        "# Iterate through the first axis of cosine_similarities_probes\n",
        "for i in range(50):\n",
        "    if optimal_threshhold < cosine_similarities_probes[i][0][0]:\n",
        "        if cosine_similarities_probes[i][0][1] == 1:\n",
        "            # Select the first element and retrieve the similarity score of that selection\n",
        "            rank1_identification_withThreshold[i] = (cosine_similarities_probes[i][0][1])\n",
        "\n",
        "    # Checking if all top 5 scores are above the optimal threshhold\n",
        "    if optimal_threshhold < cosine_similarities_probes[i][4][0]:\n",
        "        # Iterate through the second axis of cosine_similarities_probes\n",
        "        for j in range(5):\n",
        "            # Retrieve the positive label if there is one\n",
        "            if cosine_similarities_probes[i][j][1] == 1:\n",
        "                rank5_identification_withThreshold[i] = (cosine_similarities_probes[i][j][1])\n",
        "\n",
        "# Calculating the identification rates\n",
        "rank1_identification_rate = np.mean(rank1_identification)\n",
        "rank5_identification_rate = np.mean(rank5_identification)\n",
        "rank1_identification_withThreshold_rate = np.mean(rank1_identification_withThreshold)\n",
        "rank5_identification_withThreshold_rate = np.mean(rank5_identification_withThreshold)\n",
        "\n",
        "print(rank1_identification)\n",
        "print(rank5_identification)\n",
        "print(rank1_identification_withThreshold)\n",
        "print(rank5_identification_withThreshold)\n",
        "\n",
        "print(rank1_identification_rate)\n",
        "print(rank5_identification_rate)\n",
        "print(rank1_identification_withThreshold_rate)\n",
        "print(rank5_identification_withThreshold_rate)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "c-MVScEntK1p",
        "outputId": "4b1952b4-7df7-49c3-ca8a-6f85b49ade35"
      },
      "execution_count": 15,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[0. 0. 0. 0. 0. 1. 1. 1. 1. 1. 1. 0. 1. 1. 1. 1. 0. 1. 1. 1. 1. 0. 0. 0.\n",
            " 0. 1. 0. 0. 1. 0. 0. 1. 0. 0. 0. 1. 0. 0. 0. 1. 0. 0. 1. 0. 0. 1. 1. 1.\n",
            " 0. 0.]\n",
            "[0. 0. 0. 0. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 0. 1. 1. 1. 1. 0. 0. 0.\n",
            " 0. 1. 1. 0. 1. 0. 0. 1. 1. 0. 1. 1. 0. 0. 0. 1. 0. 1. 1. 1. 0. 1. 1. 1.\n",
            " 0. 1.]\n",
            "[0. 0. 0. 0. 0. 1. 1. 1. 1. 1. 1. 0. 1. 1. 1. 1. 0. 1. 1. 1. 1. 0. 0. 0.\n",
            " 0. 1. 0. 0. 1. 0. 0. 1. 0. 0. 0. 1. 0. 0. 0. 1. 0. 0. 1. 0. 0. 1. 1. 1.\n",
            " 0. 0.]\n",
            "[0. 0. 0. 0. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 0. 1. 1. 1. 1. 0. 0. 0.\n",
            " 0. 1. 1. 0. 1. 0. 0. 1. 1. 0. 1. 1. 0. 0. 0. 1. 0. 1. 1. 1. 0. 1. 1. 1.\n",
            " 0. 1.]\n",
            "0.46\n",
            "0.62\n",
            "0.46\n",
            "0.62\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "vbSQFqHYD5EQ"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Define data generators\n",
        "train_datagen = ImageDataGenerator(\n",
        "    rescale=1./255,\n",
        "    rotation_range=20,  # rotate images by up to 20 degrees\n",
        "    width_shift_range=0.2,  # shift images horizontally by up to 20% of the image width\n",
        "    height_shift_range=0.2,  # shift images vertically by up to 20% of the image height\n",
        "    shear_range=0.2,  # apply shear transformation with a shear angle of up to 20 degrees\n",
        "    zoom_range=0.2,  # apply zoom transformation by up to 20%\n",
        "    horizontal_flip=True,  # flip images horizontally\n",
        "    fill_mode='nearest'  # fill any gaps created by the above transformations with the nearest pixel\n",
        ")\n",
        "val_datagen = ImageDataGenerator(\n",
        "    rescale=1./255\n",
        ")\n",
        "\n",
        "\n",
        "train_generator = train_datagen.flow_from_directory(\n",
        "    train_dir,\n",
        "    target_size = (227,227),\n",
        "    batch_size=64,\n",
        "    class_mode='categorical')\n",
        "\n",
        "val_generator = val_datagen.flow_from_directory(\n",
        "    val_dir,\n",
        "    target_size = (227,227),\n",
        "    batch_size=64,\n",
        "    class_mode='categorical')"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "2u2MjB6FnO75",
        "outputId": "444e5eef-9ba9-43c9-db1b-ff356a4f3467"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Found 350 images belonging to 50 classes.\n",
            "Found 150 images belonging to 50 classes.\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Load VGG19 model without top layers\n",
        "base_model = VGG19(weights='imagenet', include_top=False)\n",
        "\n",
        "# Add top layers\n",
        "x = base_model.output\n",
        "x1 = GlobalAveragePooling2D()(x)\n",
        "x2 = Dense(4096, activation='relu')(x1) # Keep the fully connected layer with 4096 units\n",
        "x3 = Dense(4096, activation='relu')(x2) # Keep the fully connected layer with 4096 units\n",
        "predictions = Dense(50, activation='softmax')(x3) # Change to 50 units for 50-class classifier\n",
        "\n",
        "# Create a new model with the added top layers\n",
        "model_2 = Model(inputs=base_model.input, outputs=predictions)\n",
        "\n",
        "# Freeze base model layers\n",
        "for layer in base_model.layers:\n",
        "    layer.trainable = False\n",
        "\n",
        "# Compile the model\n",
        "model_2.compile(optimizer='adam', loss='categorical_crossentropy', metrics=['accuracy'])\n",
        "\n",
        "# Print the model summary\n",
        "model_2.summary()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "PbMt_kLvFlx7",
        "outputId": "ec362fb7-8c61-4676-83bc-653d1c255454"
      },
      "execution_count": 16,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"model_2\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " input_2 (InputLayer)        [(None, None, None, 3)]   0         \n",
            "                                                                 \n",
            " block1_conv1 (Conv2D)       (None, None, None, 64)    1792      \n",
            "                                                                 \n",
            " block1_conv2 (Conv2D)       (None, None, None, 64)    36928     \n",
            "                                                                 \n",
            " block1_pool (MaxPooling2D)  (None, None, None, 64)    0         \n",
            "                                                                 \n",
            " block2_conv1 (Conv2D)       (None, None, None, 128)   73856     \n",
            "                                                                 \n",
            " block2_conv2 (Conv2D)       (None, None, None, 128)   147584    \n",
            "                                                                 \n",
            " block2_pool (MaxPooling2D)  (None, None, None, 128)   0         \n",
            "                                                                 \n",
            " block3_conv1 (Conv2D)       (None, None, None, 256)   295168    \n",
            "                                                                 \n",
            " block3_conv2 (Conv2D)       (None, None, None, 256)   590080    \n",
            "                                                                 \n",
            " block3_conv3 (Conv2D)       (None, None, None, 256)   590080    \n",
            "                                                                 \n",
            " block3_conv4 (Conv2D)       (None, None, None, 256)   590080    \n",
            "                                                                 \n",
            " block3_pool (MaxPooling2D)  (None, None, None, 256)   0         \n",
            "                                                                 \n",
            " block4_conv1 (Conv2D)       (None, None, None, 512)   1180160   \n",
            "                                                                 \n",
            " block4_conv2 (Conv2D)       (None, None, None, 512)   2359808   \n",
            "                                                                 \n",
            " block4_conv3 (Conv2D)       (None, None, None, 512)   2359808   \n",
            "                                                                 \n",
            " block4_conv4 (Conv2D)       (None, None, None, 512)   2359808   \n",
            "                                                                 \n",
            " block4_pool (MaxPooling2D)  (None, None, None, 512)   0         \n",
            "                                                                 \n",
            " block5_conv1 (Conv2D)       (None, None, None, 512)   2359808   \n",
            "                                                                 \n",
            " block5_conv2 (Conv2D)       (None, None, None, 512)   2359808   \n",
            "                                                                 \n",
            " block5_conv3 (Conv2D)       (None, None, None, 512)   2359808   \n",
            "                                                                 \n",
            " block5_conv4 (Conv2D)       (None, None, None, 512)   2359808   \n",
            "                                                                 \n",
            " block5_pool (MaxPooling2D)  (None, None, None, 512)   0         \n",
            "                                                                 \n",
            " global_average_pooling2d_1   (None, 512)              0         \n",
            " (GlobalAveragePooling2D)                                        \n",
            "                                                                 \n",
            " dense_3 (Dense)             (None, 4096)              2101248   \n",
            "                                                                 \n",
            " dense_4 (Dense)             (None, 4096)              16781312  \n",
            "                                                                 \n",
            " dense_5 (Dense)             (None, 50)                204850    \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 39,111,794\n",
            "Trainable params: 19,087,410\n",
            "Non-trainable params: 20,024,384\n",
            "_________________________________________________________________\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Train the model\n",
        "history2_1 = model_2.fit(\n",
        "    train_generator,\n",
        "    epochs=20,  # specify the number of epochs for training\n",
        "    validation_data=val_generator\n",
        ")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "n9fF2VHcF-_S",
        "outputId": "ba255796-4667-4b7f-f9f3-c82caea469ef"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/20\n",
            "6/6 [==============================] - 361s 63s/step - loss: 4.7056 - accuracy: 0.0086 - val_loss: 3.9271 - val_accuracy: 0.0600\n",
            "Epoch 2/20\n",
            "6/6 [==============================] - 358s 63s/step - loss: 3.9366 - accuracy: 0.0171 - val_loss: 3.9015 - val_accuracy: 0.0200\n",
            "Epoch 3/20\n",
            "6/6 [==============================] - 389s 73s/step - loss: 3.9118 - accuracy: 0.0286 - val_loss: 3.8820 - val_accuracy: 0.0400\n",
            "Epoch 4/20\n",
            "6/6 [==============================] - 389s 73s/step - loss: 3.8794 - accuracy: 0.0571 - val_loss: 3.8410 - val_accuracy: 0.0733\n",
            "Epoch 5/20\n",
            "6/6 [==============================] - 352s 61s/step - loss: 3.8266 - accuracy: 0.0600 - val_loss: 3.7607 - val_accuracy: 0.0733\n",
            "Epoch 6/20\n",
            "6/6 [==============================] - 395s 70s/step - loss: 3.7113 - accuracy: 0.0686 - val_loss: 3.5992 - val_accuracy: 0.0800\n",
            "Epoch 7/20\n",
            "6/6 [==============================] - 388s 68s/step - loss: 3.5540 - accuracy: 0.1057 - val_loss: 3.4097 - val_accuracy: 0.1267\n",
            "Epoch 8/20\n",
            "6/6 [==============================] - 357s 61s/step - loss: 3.4624 - accuracy: 0.0771 - val_loss: 3.3340 - val_accuracy: 0.0533\n",
            "Epoch 9/20\n",
            "6/6 [==============================] - 388s 68s/step - loss: 3.3798 - accuracy: 0.0971 - val_loss: 3.2774 - val_accuracy: 0.1267\n",
            "Epoch 10/20\n",
            "6/6 [==============================] - 387s 68s/step - loss: 3.2321 - accuracy: 0.1229 - val_loss: 3.3238 - val_accuracy: 0.0600\n",
            "Epoch 11/20\n",
            "6/6 [==============================] - 395s 70s/step - loss: 3.1696 - accuracy: 0.1314 - val_loss: 3.2699 - val_accuracy: 0.1333\n",
            "Epoch 12/20\n",
            "6/6 [==============================] - 388s 73s/step - loss: 3.0429 - accuracy: 0.1486 - val_loss: 3.1171 - val_accuracy: 0.1200\n",
            "Epoch 13/20\n",
            "6/6 [==============================] - 352s 61s/step - loss: 2.9411 - accuracy: 0.2114 - val_loss: 3.2864 - val_accuracy: 0.0600\n",
            "Epoch 14/20\n",
            "6/6 [==============================] - 391s 69s/step - loss: 2.8392 - accuracy: 0.2057 - val_loss: 2.9638 - val_accuracy: 0.1333\n",
            "Epoch 15/20\n",
            "6/6 [==============================] - 388s 73s/step - loss: 2.8014 - accuracy: 0.1800 - val_loss: 2.9150 - val_accuracy: 0.1800\n",
            "Epoch 16/20\n",
            "6/6 [==============================] - 389s 73s/step - loss: 2.6993 - accuracy: 0.2114 - val_loss: 2.9401 - val_accuracy: 0.1667\n",
            "Epoch 17/20\n",
            "6/6 [==============================] - 349s 61s/step - loss: 2.6363 - accuracy: 0.2429 - val_loss: 2.8992 - val_accuracy: 0.1667\n",
            "Epoch 18/20\n",
            "6/6 [==============================] - 389s 68s/step - loss: 2.5746 - accuracy: 0.2629 - val_loss: 3.0145 - val_accuracy: 0.1067\n",
            "Epoch 19/20\n",
            "6/6 [==============================] - 389s 68s/step - loss: 2.4406 - accuracy: 0.2886 - val_loss: 2.6817 - val_accuracy: 0.2133\n",
            "Epoch 20/20\n",
            "6/6 [==============================] - 388s 68s/step - loss: 2.3522 - accuracy: 0.3486 - val_loss: 2.7309 - val_accuracy: 0.2000\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Plot the training and validation accuracy\n",
        "plt.plot(history2_1.history['accuracy'])\n",
        "plt.plot(history2_1.history['val_accuracy'])\n",
        "plt.title('Model Accuracy')\n",
        "plt.ylabel('Accuracy')\n",
        "plt.xlabel('Epoch')\n",
        "plt.legend(['train', 'val'], loc='upper left')\n",
        "plt.show()\n",
        "\n",
        "# Plot the training and validation loss\n",
        "plt.plot(history2_1.history['loss'])\n",
        "plt.plot(history2_1.history['val_loss'])\n",
        "plt.title('Model Loss')\n",
        "plt.ylabel('Loss')\n",
        "plt.xlabel('Epoch')\n",
        "plt.legend(['train', 'val'], loc='upper left')\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 927
        },
        "id": "--ffhsGrlleT",
        "outputId": "525d47c0-f3c8-46db-d256-ce815c8ee1d8"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "model_2.save_weights('model_2_weights.h5')"
      ],
      "metadata": {
        "id": "xcBT9quDl9of"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "model_2.save('model_2.h5')"
      ],
      "metadata": {
        "id": "BWQCJl2KmGmG"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Load the model from the H5 file\n",
        "model_2.load_weights('/content/model_2_weights.h5')"
      ],
      "metadata": {
        "id": "-gsMd2z4U1sf"
      },
      "execution_count": 17,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Create a new model that outputs the fc7 features\n",
        "fc7_layer = model_2.get_layer('dense_4')\n",
        "fc7_output = fc7_layer.output\n",
        "model_2_withoudHead = Model(inputs=model_2.input, outputs=fc7_output)"
      ],
      "metadata": {
        "id": "CFHUDTVOVOZt"
      },
      "execution_count": 27,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Define the paths to the validation dataset\n",
        "test_dir = '/content/process_data/test'\n",
        "\n",
        "# Define the batch size for validation data\n",
        "batch_size = 1\n",
        "\n",
        "# Create an empty list to hold the feature vectors for the validation images\n",
        "test_features2 = []\n",
        "\n",
        "# Loop through the validation image directories\n",
        "for folder_name in os.listdir(test_dir):\n",
        "\n",
        "    test_directory = os.path.join(test_dir, folder_name)\n",
        "\n",
        "    # Create an empty list to hold the feature vectors for the validation images in the current directory\n",
        "    directory_features = []\n",
        "    \n",
        "    # Loop through the validation images in the current directory\n",
        "    for i in range(11, 16):\n",
        "        # Load the validation image using PIL\n",
        "        img_path = os.path.join(test_directory, f'{folder_name}_{i}.jpg')\n",
        "        img = image.load_img(img_path, target_size=(227, 227))\n",
        "        \n",
        "        # Convert the image to a numpy array and preprocess it\n",
        "        x = image.img_to_array(img)\n",
        "        x = np.expand_dims(x, axis=0)\n",
        "        x = preprocess_input(x)\n",
        "        \n",
        "        # Extract the features using the VGG19 model\n",
        "        features = model_2_withoudHead.predict(x)\n",
        "        \n",
        "        # Add the features to the list of features for the current directory\n",
        "        directory_features.append(features.flatten())\n",
        "    \n",
        "    # Convert the list of features for the current directory to a numpy array\n",
        "    directory_features = np.array(directory_features)\n",
        "    \n",
        "    # Add the numpy array to the list of feature vectors for all validation images\n",
        "    test_features2.append(directory_features)\n",
        "\n",
        "# Convert the list of feature vectors for all validation images to a numpy array\n",
        "test_features2 = np.array(test_features2)\n",
        "print(test_features2)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "uElRBTuUrk0C",
        "outputId": "35ccfc28-1e7e-4b65-8451-df2bdcdd2f89"
      },
      "execution_count": 28,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 820ms/step\n",
            "1/1 [==============================] - 1s 656ms/step\n",
            "1/1 [==============================] - 1s 618ms/step\n",
            "1/1 [==============================] - 1s 637ms/step\n",
            "1/1 [==============================] - 1s 643ms/step\n",
            "1/1 [==============================] - 1s 639ms/step\n",
            "1/1 [==============================] - 1s 642ms/step\n",
            "1/1 [==============================] - 1s 648ms/step\n",
            "1/1 [==============================] - 1s 624ms/step\n",
            "1/1 [==============================] - 1s 623ms/step\n",
            "1/1 [==============================] - 1s 667ms/step\n",
            "1/1 [==============================] - 1s 611ms/step\n",
            "1/1 [==============================] - 1s 626ms/step\n",
            "1/1 [==============================] - 1s 630ms/step\n",
            "1/1 [==============================] - 1s 613ms/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 795ms/step\n",
            "1/1 [==============================] - 1s 609ms/step\n",
            "1/1 [==============================] - 1s 624ms/step\n",
            "1/1 [==============================] - 1s 607ms/step\n",
            "1/1 [==============================] - 1s 627ms/step\n",
            "1/1 [==============================] - 1s 621ms/step\n",
            "1/1 [==============================] - 1s 647ms/step\n",
            "1/1 [==============================] - 1s 622ms/step\n",
            "1/1 [==============================] - 1s 634ms/step\n",
            "1/1 [==============================] - 1s 659ms/step\n",
            "1/1 [==============================] - 1s 627ms/step\n",
            "1/1 [==============================] - 1s 638ms/step\n",
            "1/1 [==============================] - 1s 631ms/step\n",
            "1/1 [==============================] - 1s 627ms/step\n",
            "1/1 [==============================] - 1s 632ms/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 890ms/step\n",
            "1/1 [==============================] - 1s 627ms/step\n",
            "1/1 [==============================] - 1s 617ms/step\n",
            "1/1 [==============================] - 1s 619ms/step\n",
            "1/1 [==============================] - 1s 627ms/step\n",
            "1/1 [==============================] - 1s 608ms/step\n",
            "1/1 [==============================] - 1s 622ms/step\n",
            "1/1 [==============================] - 1s 621ms/step\n",
            "1/1 [==============================] - 1s 602ms/step\n",
            "1/1 [==============================] - 1s 631ms/step\n",
            "1/1 [==============================] - 1s 633ms/step\n",
            "1/1 [==============================] - 1s 623ms/step\n",
            "1/1 [==============================] - 1s 644ms/step\n",
            "1/1 [==============================] - 1s 684ms/step\n",
            "1/1 [==============================] - 1s 635ms/step\n",
            "1/1 [==============================] - 1s 967ms/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 646ms/step\n",
            "1/1 [==============================] - 1s 640ms/step\n",
            "1/1 [==============================] - 1s 638ms/step\n",
            "1/1 [==============================] - 1s 620ms/step\n",
            "1/1 [==============================] - 1s 617ms/step\n",
            "1/1 [==============================] - 1s 624ms/step\n",
            "1/1 [==============================] - 1s 615ms/step\n",
            "1/1 [==============================] - 1s 624ms/step\n",
            "1/1 [==============================] - 1s 628ms/step\n",
            "1/1 [==============================] - 1s 673ms/step\n",
            "1/1 [==============================] - 1s 682ms/step\n",
            "1/1 [==============================] - 1s 674ms/step\n",
            "1/1 [==============================] - 1s 630ms/step\n",
            "1/1 [==============================] - 1s 645ms/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 895ms/step\n",
            "1/1 [==============================] - 1s 650ms/step\n",
            "1/1 [==============================] - 1s 670ms/step\n",
            "1/1 [==============================] - 1s 632ms/step\n",
            "1/1 [==============================] - 1s 635ms/step\n",
            "1/1 [==============================] - 1s 638ms/step\n",
            "1/1 [==============================] - 1s 629ms/step\n",
            "1/1 [==============================] - 1s 631ms/step\n",
            "1/1 [==============================] - 1s 635ms/step\n",
            "1/1 [==============================] - 1s 640ms/step\n",
            "1/1 [==============================] - 1s 635ms/step\n",
            "1/1 [==============================] - 1s 648ms/step\n",
            "1/1 [==============================] - 1s 650ms/step\n",
            "1/1 [==============================] - 1s 642ms/step\n",
            "1/1 [==============================] - 1s 636ms/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 849ms/step\n",
            "1/1 [==============================] - 1s 636ms/step\n",
            "1/1 [==============================] - 1s 646ms/step\n",
            "1/1 [==============================] - 1s 651ms/step\n",
            "1/1 [==============================] - 1s 653ms/step\n",
            "1/1 [==============================] - 1s 633ms/step\n",
            "1/1 [==============================] - 1s 648ms/step\n",
            "1/1 [==============================] - 1s 652ms/step\n",
            "1/1 [==============================] - 1s 631ms/step\n",
            "1/1 [==============================] - 1s 651ms/step\n",
            "1/1 [==============================] - 1s 622ms/step\n",
            "1/1 [==============================] - 1s 633ms/step\n",
            "1/1 [==============================] - 1s 644ms/step\n",
            "1/1 [==============================] - 1s 648ms/step\n",
            "1/1 [==============================] - 1s 690ms/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 684ms/step\n",
            "1/1 [==============================] - 1s 666ms/step\n",
            "1/1 [==============================] - 1s 639ms/step\n",
            "1/1 [==============================] - 1s 654ms/step\n",
            "1/1 [==============================] - 1s 643ms/step\n",
            "1/1 [==============================] - 1s 673ms/step\n",
            "1/1 [==============================] - 1s 656ms/step\n",
            "1/1 [==============================] - 1s 639ms/step\n",
            "1/1 [==============================] - 1s 642ms/step\n",
            "1/1 [==============================] - 1s 652ms/step\n",
            "1/1 [==============================] - 1s 649ms/step\n",
            "1/1 [==============================] - 1s 634ms/step\n",
            "1/1 [==============================] - 1s 644ms/step\n",
            "1/1 [==============================] - 1s 647ms/step\n",
            "1/1 [==============================] - 1s 914ms/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 637ms/step\n",
            "1/1 [==============================] - 1s 649ms/step\n",
            "1/1 [==============================] - 1s 655ms/step\n",
            "1/1 [==============================] - 1s 670ms/step\n",
            "1/1 [==============================] - 1s 633ms/step\n",
            "1/1 [==============================] - 1s 672ms/step\n",
            "1/1 [==============================] - 1s 664ms/step\n",
            "1/1 [==============================] - 1s 641ms/step\n",
            "1/1 [==============================] - 1s 655ms/step\n",
            "1/1 [==============================] - 1s 650ms/step\n",
            "1/1 [==============================] - 1s 635ms/step\n",
            "1/1 [==============================] - 1s 659ms/step\n",
            "1/1 [==============================] - 1s 632ms/step\n",
            "1/1 [==============================] - 1s 627ms/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 984ms/step\n",
            "1/1 [==============================] - 1s 628ms/step\n",
            "1/1 [==============================] - 1s 667ms/step\n",
            "1/1 [==============================] - 1s 686ms/step\n",
            "1/1 [==============================] - 1s 689ms/step\n",
            "1/1 [==============================] - 1s 668ms/step\n",
            "1/1 [==============================] - 1s 694ms/step\n",
            "1/1 [==============================] - 1s 668ms/step\n",
            "1/1 [==============================] - 1s 658ms/step\n",
            "1/1 [==============================] - 1s 682ms/step\n",
            "1/1 [==============================] - 1s 676ms/step\n",
            "1/1 [==============================] - 1s 639ms/step\n",
            "1/1 [==============================] - 1s 658ms/step\n",
            "1/1 [==============================] - 1s 699ms/step\n",
            "1/1 [==============================] - 1s 975ms/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 691ms/step\n",
            "1/1 [==============================] - 1s 660ms/step\n",
            "1/1 [==============================] - 1s 639ms/step\n",
            "1/1 [==============================] - 1s 668ms/step\n",
            "1/1 [==============================] - 1s 679ms/step\n",
            "1/1 [==============================] - 1s 644ms/step\n",
            "1/1 [==============================] - 1s 638ms/step\n",
            "1/1 [==============================] - 1s 668ms/step\n",
            "1/1 [==============================] - 1s 648ms/step\n",
            "1/1 [==============================] - 1s 630ms/step\n",
            "1/1 [==============================] - 1s 649ms/step\n",
            "1/1 [==============================] - 1s 627ms/step\n",
            "1/1 [==============================] - 1s 673ms/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 818ms/step\n",
            "1/1 [==============================] - 1s 622ms/step\n",
            "1/1 [==============================] - 1s 658ms/step\n",
            "1/1 [==============================] - 1s 657ms/step\n",
            "1/1 [==============================] - 1s 657ms/step\n",
            "1/1 [==============================] - 1s 641ms/step\n",
            "1/1 [==============================] - 1s 659ms/step\n",
            "1/1 [==============================] - 1s 649ms/step\n",
            "1/1 [==============================] - 1s 649ms/step\n",
            "1/1 [==============================] - 1s 664ms/step\n",
            "1/1 [==============================] - 1s 634ms/step\n",
            "1/1 [==============================] - 1s 650ms/step\n",
            "1/1 [==============================] - 1s 671ms/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 784ms/step\n",
            "1/1 [==============================] - 1s 621ms/step\n",
            "1/1 [==============================] - 1s 645ms/step\n",
            "1/1 [==============================] - 1s 654ms/step\n",
            "1/1 [==============================] - 1s 643ms/step\n",
            "1/1 [==============================] - 1s 647ms/step\n",
            "1/1 [==============================] - 1s 635ms/step\n",
            "1/1 [==============================] - 1s 652ms/step\n",
            "1/1 [==============================] - 1s 650ms/step\n",
            "1/1 [==============================] - 1s 645ms/step\n",
            "1/1 [==============================] - 1s 651ms/step\n",
            "1/1 [==============================] - 1s 666ms/step\n",
            "1/1 [==============================] - 1s 631ms/step\n",
            "1/1 [==============================] - 1s 750ms/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 735ms/step\n",
            "1/1 [==============================] - 1s 627ms/step\n",
            "1/1 [==============================] - 1s 624ms/step\n",
            "1/1 [==============================] - 1s 612ms/step\n",
            "1/1 [==============================] - 1s 640ms/step\n",
            "1/1 [==============================] - 1s 633ms/step\n",
            "1/1 [==============================] - 1s 617ms/step\n",
            "1/1 [==============================] - 1s 626ms/step\n",
            "1/1 [==============================] - 1s 636ms/step\n",
            "1/1 [==============================] - 1s 610ms/step\n",
            "1/1 [==============================] - 1s 636ms/step\n",
            "1/1 [==============================] - 1s 636ms/step\n",
            "1/1 [==============================] - 1s 640ms/step\n",
            "1/1 [==============================] - 1s 730ms/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 908ms/step\n",
            "1/1 [==============================] - 1s 637ms/step\n",
            "1/1 [==============================] - 1s 635ms/step\n",
            "1/1 [==============================] - 1s 652ms/step\n",
            "1/1 [==============================] - 1s 673ms/step\n",
            "1/1 [==============================] - 1s 673ms/step\n",
            "1/1 [==============================] - 1s 672ms/step\n",
            "1/1 [==============================] - 1s 626ms/step\n",
            "1/1 [==============================] - 1s 642ms/step\n",
            "1/1 [==============================] - 1s 662ms/step\n",
            "1/1 [==============================] - 1s 629ms/step\n",
            "1/1 [==============================] - 1s 635ms/step\n",
            "1/1 [==============================] - 1s 633ms/step\n",
            "1/1 [==============================] - 1s 657ms/step\n",
            "1/1 [==============================] - 1s 758ms/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 2s 2s/step\n",
            "1/1 [==============================] - 2s 2s/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 651ms/step\n",
            "1/1 [==============================] - 1s 630ms/step\n",
            "1/1 [==============================] - 1s 621ms/step\n",
            "1/1 [==============================] - 1s 621ms/step\n",
            "1/1 [==============================] - 1s 659ms/step\n",
            "1/1 [==============================] - 1s 945ms/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 665ms/step\n",
            "1/1 [==============================] - 1s 980ms/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 641ms/step\n",
            "1/1 [==============================] - 1s 615ms/step\n",
            "1/1 [==============================] - 1s 667ms/step\n",
            "1/1 [==============================] - 1s 654ms/step\n",
            "[[[0.         0.         0.         ... 0.         0.         0.        ]\n",
            "  [0.         0.         0.         ... 0.         0.         0.        ]\n",
            "  [0.         0.36846513 0.         ... 0.         0.         0.        ]\n",
            "  [0.         0.         0.         ... 0.         0.         0.        ]\n",
            "  [0.         0.         0.         ... 0.         0.         0.        ]]\n",
            "\n",
            " [[0.         0.         0.         ... 0.         0.         0.        ]\n",
            "  [0.         0.         0.         ... 0.         0.         0.        ]\n",
            "  [0.         0.90180546 0.         ... 0.         0.         0.        ]\n",
            "  [0.         0.6866916  0.         ... 0.         0.         0.        ]\n",
            "  [0.         0.         0.         ... 0.         0.         0.        ]]\n",
            "\n",
            " [[0.         0.         0.         ... 0.         0.         0.        ]\n",
            "  [0.         1.8151821  0.         ... 0.         0.         0.        ]\n",
            "  [0.         0.9168182  0.         ... 0.         0.         0.        ]\n",
            "  [0.         0.7471268  0.         ... 0.         0.         0.        ]\n",
            "  [0.         0.9814264  0.         ... 0.         0.         0.        ]]\n",
            "\n",
            " ...\n",
            "\n",
            " [[0.         0.         0.         ... 0.         0.         0.        ]\n",
            "  [0.         0.04693278 0.         ... 0.         0.         0.        ]\n",
            "  [0.         0.28415695 0.         ... 0.         0.         0.        ]\n",
            "  [0.         0.         0.         ... 0.         0.         0.        ]\n",
            "  [0.         0.         0.         ... 0.         0.         0.        ]]\n",
            "\n",
            " [[0.         0.30285445 0.         ... 0.         0.         0.        ]\n",
            "  [0.         0.         0.         ... 0.         0.         0.        ]\n",
            "  [0.         1.6949345  0.         ... 0.         0.         0.        ]\n",
            "  [0.         2.0286076  0.         ... 0.         0.         0.        ]\n",
            "  [0.         2.0377614  0.         ... 0.         0.         0.        ]]\n",
            "\n",
            " [[0.         0.         0.         ... 0.         0.         0.        ]\n",
            "  [0.         0.8517284  0.         ... 0.         0.         0.        ]\n",
            "  [0.         0.         0.         ... 0.         0.         0.        ]\n",
            "  [0.         0.         0.         ... 0.         0.         0.        ]\n",
            "  [0.         0.         0.         ... 0.         0.         0.        ]]]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Define data generators\n",
        "train_datagen = ImageDataGenerator(\n",
        "    rescale=1./255\n",
        ")\n",
        "val_datagen = ImageDataGenerator(\n",
        "    rescale=1./255\n",
        ")\n",
        "\n",
        "\n",
        "train_generator = train_datagen.flow_from_directory(\n",
        "    train_dir,\n",
        "    target_size = (227,227),\n",
        "    batch_size=64,\n",
        "    class_mode='categorical')\n",
        "\n",
        "val_generator = val_datagen.flow_from_directory(\n",
        "    val_dir,\n",
        "    target_size = (227,227),\n",
        "    batch_size=64,\n",
        "    class_mode='categorical')"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "UtkL-X8lJFQc",
        "outputId": "40ba7dd4-f862-421d-c36e-b52c14819f47"
      },
      "execution_count": 18,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Found 350 images belonging to 50 classes.\n",
            "Found 150 images belonging to 50 classes.\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Load VGG19 model without top layers\n",
        "base_model = VGG19(weights='imagenet', include_top=False)\n",
        "\n",
        "# Add top layers\n",
        "x = base_model.output\n",
        "x1 = GlobalAveragePooling2D()(x)\n",
        "x2 = Dense(4096, activation='relu')(x1) # Keep the fully connected layer with 4096 units\n",
        "x3 = Dense(4096, activation='relu')(x2) # Keep the fully connected layer with 4096 units\n",
        "predictions = Dense(50, activation='softmax')(x3) # Change to 50 units for 50-class classifier\n",
        "\n",
        "# Create a new model with the added top layers\n",
        "model_3 = Model(inputs=base_model.input, outputs=predictions)\n",
        "\n",
        "# Freeze base model layers\n",
        "for layer in base_model.layers:\n",
        "    layer.trainable = False\n",
        "\n",
        "# Compile the model\n",
        "model_3.compile(optimizer='adam', loss='categorical_crossentropy', metrics=['accuracy'])\n",
        "\n",
        "# Print the model summary\n",
        "model_3.summary()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "uf9-kSjVJt4f",
        "outputId": "c1473ba5-78b9-4237-f7a6-18fff0647f6c"
      },
      "execution_count": 19,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"model_3\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " input_3 (InputLayer)        [(None, None, None, 3)]   0         \n",
            "                                                                 \n",
            " block1_conv1 (Conv2D)       (None, None, None, 64)    1792      \n",
            "                                                                 \n",
            " block1_conv2 (Conv2D)       (None, None, None, 64)    36928     \n",
            "                                                                 \n",
            " block1_pool (MaxPooling2D)  (None, None, None, 64)    0         \n",
            "                                                                 \n",
            " block2_conv1 (Conv2D)       (None, None, None, 128)   73856     \n",
            "                                                                 \n",
            " block2_conv2 (Conv2D)       (None, None, None, 128)   147584    \n",
            "                                                                 \n",
            " block2_pool (MaxPooling2D)  (None, None, None, 128)   0         \n",
            "                                                                 \n",
            " block3_conv1 (Conv2D)       (None, None, None, 256)   295168    \n",
            "                                                                 \n",
            " block3_conv2 (Conv2D)       (None, None, None, 256)   590080    \n",
            "                                                                 \n",
            " block3_conv3 (Conv2D)       (None, None, None, 256)   590080    \n",
            "                                                                 \n",
            " block3_conv4 (Conv2D)       (None, None, None, 256)   590080    \n",
            "                                                                 \n",
            " block3_pool (MaxPooling2D)  (None, None, None, 256)   0         \n",
            "                                                                 \n",
            " block4_conv1 (Conv2D)       (None, None, None, 512)   1180160   \n",
            "                                                                 \n",
            " block4_conv2 (Conv2D)       (None, None, None, 512)   2359808   \n",
            "                                                                 \n",
            " block4_conv3 (Conv2D)       (None, None, None, 512)   2359808   \n",
            "                                                                 \n",
            " block4_conv4 (Conv2D)       (None, None, None, 512)   2359808   \n",
            "                                                                 \n",
            " block4_pool (MaxPooling2D)  (None, None, None, 512)   0         \n",
            "                                                                 \n",
            " block5_conv1 (Conv2D)       (None, None, None, 512)   2359808   \n",
            "                                                                 \n",
            " block5_conv2 (Conv2D)       (None, None, None, 512)   2359808   \n",
            "                                                                 \n",
            " block5_conv3 (Conv2D)       (None, None, None, 512)   2359808   \n",
            "                                                                 \n",
            " block5_conv4 (Conv2D)       (None, None, None, 512)   2359808   \n",
            "                                                                 \n",
            " block5_pool (MaxPooling2D)  (None, None, None, 512)   0         \n",
            "                                                                 \n",
            " global_average_pooling2d_2   (None, 512)              0         \n",
            " (GlobalAveragePooling2D)                                        \n",
            "                                                                 \n",
            " dense_6 (Dense)             (None, 4096)              2101248   \n",
            "                                                                 \n",
            " dense_7 (Dense)             (None, 4096)              16781312  \n",
            "                                                                 \n",
            " dense_8 (Dense)             (None, 50)                204850    \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 39,111,794\n",
            "Trainable params: 19,087,410\n",
            "Non-trainable params: 20,024,384\n",
            "_________________________________________________________________\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Train the model\n",
        "history_3_1 = model_3.fit(\n",
        "    train_generator,\n",
        "    epochs=20,  # specify the number of epochs for training\n",
        "    validation_data=val_generator\n",
        ")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "7QMtYCHqJ61Z",
        "outputId": "c6ee5067-7bcc-4239-aa60-c659f98a21f8"
      },
      "execution_count": 20,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/20\n",
            "6/6 [==============================] - 387s 68s/step - loss: 4.5408 - accuracy: 0.0114 - val_loss: 3.9132 - val_accuracy: 0.0333\n",
            "Epoch 2/20\n",
            "6/6 [==============================] - 385s 68s/step - loss: 3.9181 - accuracy: 0.0257 - val_loss: 3.8848 - val_accuracy: 0.0467\n",
            "Epoch 3/20\n",
            "6/6 [==============================] - 379s 67s/step - loss: 3.8693 - accuracy: 0.0629 - val_loss: 3.8268 - val_accuracy: 0.1000\n",
            "Epoch 4/20\n",
            "6/6 [==============================] - 385s 67s/step - loss: 3.7731 - accuracy: 0.0886 - val_loss: 3.7000 - val_accuracy: 0.0800\n",
            "Epoch 5/20\n",
            "6/6 [==============================] - 380s 67s/step - loss: 3.5886 - accuracy: 0.1114 - val_loss: 3.5037 - val_accuracy: 0.1133\n",
            "Epoch 6/20\n",
            "6/6 [==============================] - 336s 63s/step - loss: 3.3741 - accuracy: 0.1229 - val_loss: 3.3142 - val_accuracy: 0.1267\n",
            "Epoch 7/20\n",
            "6/6 [==============================] - 377s 71s/step - loss: 3.1539 - accuracy: 0.1400 - val_loss: 3.2029 - val_accuracy: 0.1933\n",
            "Epoch 8/20\n",
            "6/6 [==============================] - 380s 72s/step - loss: 2.9942 - accuracy: 0.2114 - val_loss: 3.0430 - val_accuracy: 0.1800\n",
            "Epoch 9/20\n",
            "6/6 [==============================] - 377s 71s/step - loss: 2.8280 - accuracy: 0.2114 - val_loss: 3.0120 - val_accuracy: 0.2067\n",
            "Epoch 10/20\n",
            "6/6 [==============================] - 378s 67s/step - loss: 2.5934 - accuracy: 0.3400 - val_loss: 2.9660 - val_accuracy: 0.1800\n",
            "Epoch 11/20\n",
            "6/6 [==============================] - 383s 68s/step - loss: 2.4251 - accuracy: 0.3200 - val_loss: 2.8124 - val_accuracy: 0.2000\n",
            "Epoch 12/20\n",
            "6/6 [==============================] - 341s 59s/step - loss: 2.2480 - accuracy: 0.3914 - val_loss: 2.7900 - val_accuracy: 0.2200\n",
            "Epoch 13/20\n",
            "6/6 [==============================] - 382s 68s/step - loss: 2.1047 - accuracy: 0.4200 - val_loss: 2.6399 - val_accuracy: 0.2667\n",
            "Epoch 14/20\n",
            "6/6 [==============================] - 405s 71s/step - loss: 1.9086 - accuracy: 0.4743 - val_loss: 2.5645 - val_accuracy: 0.2533\n",
            "Epoch 15/20\n",
            "6/6 [==============================] - 411s 71s/step - loss: 1.8393 - accuracy: 0.4857 - val_loss: 2.5064 - val_accuracy: 0.3133\n",
            "Epoch 16/20\n",
            "6/6 [==============================] - 402s 71s/step - loss: 1.7617 - accuracy: 0.5029 - val_loss: 2.5050 - val_accuracy: 0.3333\n",
            "Epoch 17/20\n",
            "6/6 [==============================] - 381s 67s/step - loss: 1.5579 - accuracy: 0.5543 - val_loss: 2.3949 - val_accuracy: 0.3000\n",
            "Epoch 18/20\n",
            "6/6 [==============================] - 389s 74s/step - loss: 1.3797 - accuracy: 0.6486 - val_loss: 2.2918 - val_accuracy: 0.4200\n",
            "Epoch 19/20\n",
            "6/6 [==============================] - 342s 60s/step - loss: 1.2344 - accuracy: 0.6943 - val_loss: 2.0652 - val_accuracy: 0.3800\n",
            "Epoch 20/20\n",
            "6/6 [==============================] - 386s 68s/step - loss: 1.1170 - accuracy: 0.7429 - val_loss: 1.9813 - val_accuracy: 0.3867\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Plot the training and validation accuracy\n",
        "plt.plot(history_3_1.history['accuracy'])\n",
        "plt.plot(history_3_1.history['val_accuracy'])\n",
        "plt.title('Model Accuracy')\n",
        "plt.ylabel('Accuracy')\n",
        "plt.xlabel('Epoch')\n",
        "plt.legend(['train', 'val'], loc='upper left')\n",
        "plt.show()\n",
        "\n",
        "# Plot the training and validation loss\n",
        "plt.plot(history_3_1.history['loss'])\n",
        "plt.plot(history_3_1.history['val_loss'])\n",
        "plt.title('Model Loss')\n",
        "plt.ylabel('Loss')\n",
        "plt.xlabel('Epoch')\n",
        "plt.legend(['train', 'val'], loc='upper left')\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 927
        },
        "id": "yjOjlMdWKQEe",
        "outputId": "538c0573-c622-4e6c-90f7-58c7a761644a"
      },
      "execution_count": 21,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAjcAAAHHCAYAAABDUnkqAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAABqb0lEQVR4nO3dd3wUdf7H8dduek8oKZBQpLeEDsGCUqRJUVApHmLBAqjYzuP82c7z0LNXBBXwVFABKaKIgPQuHaS3UBJCSydtd35/LEQDJCSQZHaT9/PxmAfs7HdmP5Mx7pvvfGe+FsMwDERERETKCavZBYiIiIiUJIUbERERKVcUbkRERKRcUbgRERGRckXhRkRERMoVhRsREREpVxRuREREpFxRuBEREZFyReFGREREyhWFGxFxShaLhZdffrnY2x06dAiLxcLkyZNLvCYRcQ0KNyJSoMmTJ2OxWLBYLKxYseKS9w3DICoqCovFwm233WZChVdvyZIlWCwWpk+fbnYpIlLCFG5E5Iq8vb2ZMmXKJeuXLl3K0aNH8fLyMqEqEZHLU7gRkSvq2bMn06ZNIzc3N9/6KVOm0KpVK8LDw02qTETkUgo3InJFgwYN4vTp0yxYsCBvXXZ2NtOnT2fw4MGX3SY9PZ2nn36aqKgovLy8aNCgAW+99RaGYeRrl5WVxZNPPknVqlUJCAigT58+HD169LL7PHbsGPfffz9hYWF4eXnRpEkTJk6cWHIHehkHDhzgzjvvpFKlSvj6+tK+fXt++umnS9p9+OGHNGnSBF9fX0JCQmjdunW+3q7U1FRGjx5NrVq18PLyIjQ0lK5du7Jx48ZSrV+kIlK4EZErqlWrFrGxsUydOjVv3bx580hOTmbgwIGXtDcMgz59+vDuu+/SvXt33nnnHRo0aMCzzz7LU089la/tgw8+yHvvvcett97K66+/joeHB7169bpknydOnKB9+/YsXLiQUaNG8f7771O3bl0eeOAB3nvvvRI/5guf2aFDB+bPn8+IESN47bXXyMzMpE+fPsycOTOv3Weffcbjjz9O48aNee+993jllVdo3rw5a9euzWvzyCOPMG7cOPr3788nn3zCM888g4+PDzt37iyV2kUqNENEpACTJk0yAGP9+vXGRx99ZAQEBBgZGRmGYRjGnXfeadxyyy2GYRhGzZo1jV69euVtN2vWLAMw/v3vf+fb34ABAwyLxWLs27fPMAzD2Lx5swEYI0aMyNdu8ODBBmC89NJLeeseeOABIyIiwjh16lS+tgMHDjSCgoLy6jp48KABGJMmTSr02BYvXmwAxrRp0wpsM3r0aAMwli9fnrcuNTXVqF27tlGrVi3DZrMZhmEYffv2NZo0aVLo5wUFBRkjR44stI2IlAz13IhIkdx1112cO3eOuXPnkpqayty5cwu8JPXzzz/j5ubG448/nm/9008/jWEYzJs3L68dcEm70aNH53ttGAYzZsygd+/eGIbBqVOn8pZu3bqRnJxcKpd3fv75Z9q2bcsNN9yQt87f35+HHnqIQ4cO8ccffwAQHBzM0aNHWb9+fYH7Cg4OZu3atRw/frzE6xSR/BRuRKRIqlatSpcuXZgyZQo//PADNpuNAQMGXLbt4cOHqVatGgEBAfnWN2rUKO/9C39arVbq1KmTr12DBg3yvT558iRJSUlMmDCBqlWr5lvuu+8+ABITE0vkOC8+jotrudxxPPfcc/j7+9O2bVvq1avHyJEjWblyZb5t/vvf/7J9+3aioqJo27YtL7/8MgcOHCjxmkUE3M0uQERcx+DBgxk+fDgJCQn06NGD4ODgMvlcu90OwD333MO999572TbR0dFlUsvlNGrUiN27dzN37lx++eUXZsyYwSeffMKLL77IK6+8Ajh6vm688UZmzpzJr7/+yptvvskbb7zBDz/8QI8ePUyrXaQ8Us+NiBTZ7bffjtVqZc2aNQVekgKoWbMmx48fJzU1Nd/6Xbt25b1/4U+73c7+/fvztdu9e3e+1xfupLLZbHTp0uWyS2hoaEkc4iXHcXEtlzsOAD8/P+6++24mTZpEXFwcvXr1yhuAfEFERAQjRoxg1qxZHDx4kMqVK/Paa6+VeN0iFZ3CjYgUmb+/P+PGjePll1+md+/eBbbr2bMnNpuNjz76KN/6d999F4vFktdTceHPDz74IF+7i+9+cnNzo3///syYMYPt27df8nknT568msO5op49e7Ju3TpWr16dty49PZ0JEyZQq1YtGjduDMDp06fzbefp6Unjxo0xDIOcnBxsNhvJycn52oSGhlKtWjWysrJKpXaRikyXpUSkWAq6LPRXvXv35pZbbuH555/n0KFDxMTE8OuvvzJ79mxGjx6dN8amefPmDBo0iE8++YTk5GQ6dOjAokWL2Ldv3yX7fP3111m8eDHt2rVj+PDhNG7cmDNnzrBx40YWLlzImTNnrup4ZsyYkdcTc/Fx/uMf/2Dq1Kn06NGDxx9/nEqVKvHll19y8OBBZsyYgdXq+PfhrbfeSnh4ONdffz1hYWHs3LmTjz76iF69ehEQEEBSUhKRkZEMGDCAmJgY/P39WbhwIevXr+ftt9++qrpFpBDm3qwlIs7sr7eCF+biW8ENw3HL9JNPPmlUq1bN8PDwMOrVq2e8+eabht1uz9fu3LlzxuOPP25UrlzZ8PPzM3r37m0cOXLkklvBDcMwTpw4YYwcOdKIiooyPDw8jPDwcKNz587GhAkT8toU91bwgpYLt3/v37/fGDBggBEcHGx4e3sbbdu2NebOnZtvX+PHjzduuukmo3LlyoaXl5dRp04d49lnnzWSk5MNwzCMrKws49lnnzViYmKMgIAAw8/Pz4iJiTE++eSTQmsUkatjMYyLHhcqIiIi4sI05kZERETKFYUbERERKVcUbkRERKRcUbgRERGRckXhRkRERMoVhRsREREpVyrcQ/zsdjvHjx8nICAAi8VidjkiIiJSBIZhkJqaSrVq1fIeoFmQChdujh8/TlRUlNlliIiIyFU4cuQIkZGRhbapcOEmICAAcPxwAgMDTa5GREREiiIlJYWoqKi87/HCVLhwc+FSVGBgoMKNiIiIiynKkBINKBYREZFyReFGREREyhWFGxERESlXKtyYm6Ky2Wzk5OSYXYZL8vDwwM3NzewyRESkglK4uYhhGCQkJJCUlGR2KS4tODiY8PBwPUtIRETKnMLNRS4Em9DQUHx9ffXlXEyGYZCRkUFiYiIAERERJlckIiIVjcLNX9hstrxgU7lyZbPLcVk+Pj4AJCYmEhoaqktUIiJSpjSg+C8ujLHx9fU1uRLXd+FnqHFLIiJS1hRuLkOXoq6dfoYiImIWhRsREREpVxRu5BK1atXivffeM7sMERGRq6IBxeXEzTffTPPmzUsklKxfvx4/P79rL0pERMQECjclKMdmJ9dm4OPpfHcHGYaBzWbD3f3Kp7xq1aplUJGIiEjp0GWpEpJ8Lptd8akcSzpX5p89bNgwli5dyvvvv4/FYsFisTB58mQsFgvz5s2jVatWeHl5sWLFCvbv30/fvn0JCwvD39+fNm3asHDhwnz7u/iylMVi4fPPP+f222/H19eXevXqMWfOnDI+ShERkaJxmnDz+uuvY7FYGD16dIFtLnxh/3Xx9vYu1boMwyAjO/eKC0Bmjo0z6VmcSc8q0jZXWgzDKFKN77//PrGxsQwfPpz4+Hji4+OJiooC4B//+Aevv/46O3fuJDo6mrS0NHr27MmiRYvYtGkT3bt3p3fv3sTFxRX6Ga+88gp33XUXW7dupWfPngwZMoQzZ85c2w9XRESkFDjFZan169czfvx4oqOjr9g2MDCQ3bt3570u7VuOz+XYaPzi/FL9jIL88a9u+Hpe+RQFBQXh6emJr68v4eHhAOzatQuAf/3rX3Tt2jWvbaVKlYiJicl7/eqrrzJz5kzmzJnDqFGjCvyMYcOGMWjQIAD+85//8MEHH7Bu3Tq6d+9+VccmIiJSWkzvuUlLS2PIkCF89tlnhISEXLG9xWIhPDw8bwkLCyuDKl1X69at871OS0vjmWeeoVGjRgQHB+Pv78/OnTuv2HPz1+Dp5+dHYGBg3hQLIiIizsT0npuRI0fSq1cvunTpwr///e8rtk9LS6NmzZrY7XZatmzJf/7zH5o0aVJg+6ysLLKysvJep6SkFKs+Hw83/vhXtyK1NQyDPSfSyLHZiQrxJcjXo1ifdbnPvlYX3/X0zDPPsGDBAt566y3q1q2Lj48PAwYMIDs7u9D9eHjkPxaLxYLdbr/m+kREREqaqeHm22+/ZePGjaxfv75I7Rs0aMDEiROJjo4mOTmZt956iw4dOrBjxw4iIyMvu83YsWN55ZVXrrpGi8VSpEtDF0QEeZOYmkVmrp2IYmx3rTw9PbHZbFdst3LlSoYNG8btt98OOMLioUOHSrk6ERGRsmPaZakjR47wxBNP8M033xR5UHBsbCxDhw6lefPmdOzYkR9++IGqVasyfvz4ArcZM2YMycnJecuRI0dK6hAuK8TPE4DUzByyc8uuZ6NWrVqsXbuWQ4cOcerUqQJ7VerVq8cPP/zA5s2b2bJlC4MHD1YPjIiIlCumhZsNGzaQmJhIy5YtcXd3x93dnaVLl/LBBx/g7u5epF4IDw8PWrRowb59+wps4+XlRWBgYL6lNHm5u+Hv5eixOZtR+KWekvTMM8/g5uZG48aNqVq1aoFjaN555x1CQkLo0KEDvXv3plu3brRs2bLM6hQRESltFqOo9xuXsNTUVA4fPpxv3X333UfDhg157rnnaNq06RX3YbPZaNKkCT179uSdd94p0uempKQQFBREcnLyJUEnMzOTgwcPUrt27Wu6xTwpI5u4Mxl4ullpEB5QISeRLKmfpYiICBT+/X0x08bcBAQEXBJg/Pz8qFy5ct76oUOHUr16dcaOHQs4bmtu3749devWJSkpiTfffJPDhw/z4IMPlnn9hQn09sDNaiHbZictK5cA72sbWCwiIiJFZ/rdUoWJi4vDav3zytnZs2cZPnw4CQkJhISE0KpVK1atWkXjxo1NrPJSVquFYF9PTqdlcSY9W+FGRESkDJl2WcosZXFZCuBcdi57E9OwWCw0Cg/A3c30RwqVKV2WEhGRklScy1IV6xu3DPl4uuPj4YZhGJzNyDG7HBERkQpD4aYUVTp/W/jZjOwizxMlIiIi10bhphQF+XpgtVjIzLFxLvvKt7aLiIjItVO4KUXuVitBPo7BxGfK8Jk3IiIiFZnCTSm78MTipIwcbHZdmhIRESltCjelzM/TDS93K3bDIPmcBhaLiIiUNoWbUmaxWPJ6b86mO++lqVq1avHee++ZXYaIiMg1U7gpAyG+nliwkJ6dS2aOBhaLiIiUJoWbMuDhZiXAu+wn0xQREamIFG7KSN4zb9JzsJfwM28mTJhAtWrVsNvt+db37duX+++/n/3799O3b1/CwsLw9/enTZs2LFy4sERrEBERcRYKN1diGJCdfs1LgDULD3smtqw0UlOSirZdEUPQnXfeyenTp1m8eHHeujNnzvDLL78wZMgQ0tLS6NmzJ4sWLWLTpk10796d3r17ExcXV0o/NBEREfM49cSZTiEnA/5T7Zp3YwEaFXejfx4HT78rNgsJCaFHjx5MmTKFzp07AzB9+nSqVKnCLbfcgtVqJSYmJq/9q6++ysyZM5kzZw6jRo0qblUiIiJOTT035cSQIUOYMWMGWVlZAHzzzTcMHDgQq9VKWloazzzzDI0aNSI4OBh/f3927typnhsRESmX1HNzJR6+jh6UEnLgVDrpWbmEBnoRFnCF2bI9fIu83969e2MYBj/99BNt2rRh+fLlvPvuuwA888wzLFiwgLfeeou6devi4+PDgAEDyM7W4GYRESl/FG6uxGIp0qWhogoJ8iDtbAZnc6yEevhisVhKZL/e3t7ccccdfPPNN+zbt48GDRrQsmVLAFauXMmwYcO4/fbbAUhLS+PQoUMl8rkiIiLORuGmjAX5eHA8yUJ2rp30rFz8vT1KbN9DhgzhtttuY8eOHdxzzz156+vVq8cPP/xA7969sVgsvPDCC5fcWSUiIlJeaMxNGbNaLQT7np9MM71kp2Po1KkTlSpVYvfu3QwePDhv/TvvvENISAgdOnSgd+/edOvWLa9XR0REpLxRz40JQvw8OZ2eTXJmDrk2O+5uJZMxrVYrx49fOj6oVq1a/Pbbb/nWjRw5Mt9rXaYSEZHyQj03JvDxcMPHww3DMEjSZJoiIiIlSuHGBH+dTPNMejZGCT+xWEREpCJTuDFJsI8HFouFzBwb5zSZpoiISIlRuDGJu5uVIB/HwOKz6XrejIiISElRuLmMsrpMVOn8XVNJGTnY7eXr0pQutYmIiFkUbv7Cw8MRNjIyMsrk8/y83PF0t2IzDJLL2cDiCz/DCz9TERGRsqJbwf/Czc2N4OBgEhMTAfD1LbknCBckwN3gVGY2J5Nt+LiV3JOQzWIYBhkZGSQmJhIcHIybm5vZJYmISAWjcHOR8PBwgLyAU9psdoOTyZkYQNYZrxJ75o3ZgoOD836WIiIiZUnh5iIWi4WIiAhCQ0PJySmbS0UTZ25j7YHT3N2mBg/dVLtMPrM0eXh4qMdGRERMo3BTADc3tzL7gr61WRQ/bEnkf+uO82jnhniUk94bERERM+hb1Al0bhRKFX9PTqVlsXhX2VwOExERKa8UbpyAh5uV/i0jAfhu/RGTqxEREXFtCjdO4q42UQAs3p3IiZRMk6sRERFxXQo3TqJOVX/a1ArBbsD0DUfNLkdERMRlKdw4kbtaO3pvvv/9SLl7YrGIiEhZUbhxIr2iI/D3cufw6QzWHjxjdjkiIiIuSeHGifh6utM7phoA362PM7kaERER16Rw42QGnh9YPG97Qrmbb0pERKQsKNw4mejIIBqGB5CVa2fO5mNmlyMiIuJyFG6cjMViyRtY/K2eeSMiIlJsThNuXn/9dSwWC6NHjy603bRp02jYsCHe3t40a9aMn3/+uWwKLEO3t6iOp5uVHcdT2H4s2exyREREXIpThJv169czfvx4oqOjC223atUqBg0axAMPPMCmTZvo168f/fr1Y/v27WVUadkI8fPk1iZhgJ5YLCIiUlymh5u0tDSGDBnCZ599RkhISKFt33//fbp3786zzz5Lo0aNePXVV2nZsiUfffRRGVVbdga2qQHArM3HyMyxmVyNiIiI6zA93IwcOZJevXrRpUuXK7ZdvXr1Je26devG6tWrC9wmKyuLlJSUfIsr6FCnMpEhPqRm5vLL9gSzyxEREXEZpoabb7/9lo0bNzJ27NgitU9ISCAsLCzfurCwMBISCv7yHzt2LEFBQXlLVFTUNdVcVqxWC3e2ujCwWM+8ERERKSrTws2RI0d44okn+Oabb/D29i61zxkzZgzJycl5y5EjrjOG5c7WkVgssObAGQ6dSje7HBEREZdgWrjZsGEDiYmJtGzZEnd3d9zd3Vm6dCkffPAB7u7u2GyXjjMJDw/nxIkT+dadOHGC8PDwAj/Hy8uLwMDAfIurqBbsw031qgKO+aZERETkykwLN507d2bbtm1s3rw5b2ndujVDhgxh8+bNuLm5XbJNbGwsixYtyrduwYIFxMbGllXZZe7CE4unbzhKrs1ucjUiIiLOz92sDw4ICKBp06b51vn5+VG5cuW89UOHDqV69ep5Y3KeeOIJOnbsyNtvv02vXr349ttv+f3335kwYUKZ119WOjcKo7KfJ4mpWSzdc5LOjcKuvJGIiEgFZvrdUoWJi4sjPj4+73WHDh2YMmUKEyZMICYmhunTpzNr1qxLQlJ54ulu5Y6W1QE9sVhERKQoLIZhGGYXUZZSUlIICgoiOTnZZcbf7EtMpcs7y3CzWlg9phOhAaU3AFtERMQZFef727TLUuVO2klY+gYERZ5fohx/BoSD9dLxQ8VRNzSAljWC2RiXxIBxq2lVM4Rm1YNoFhlEk2qB+HrqNIqIiFygb8WScuYArP/s0vVWdwioBsFRfwk+fwk/QZHgFXDF3Q+/8TpGTtlI3JkM4s5kMHOTY8ZwqwXqhvrTtHoQ0ecDT+OIIHw8ry1QiYiIuCpdliopp/fDpq8h+eifS8oxMIowdYJ3UP6wc3H48Q8HN3dOpWWx7WgyW48ms+1YMtuOJXEiJeuS3VktUC80gGaRQURHBtG0ehCNIwLx9lDgERER11Sc72+Fm9Jkt0FqwvmwcyR/8LmwLjPpyvuxuEFgNUfgqd4SanaAGrHgW4nElEy2Hfsz8Gw9msyptEsDj5vVQr1Qf6Ijg85f0gqmYXiAAo+IiLgEhZtCON2A4qxUSD5WQAA64uj9sedeftuqjaBmLNTo4Ag8QdUxDIMTKVmOnp2jSWw9lsy2o8mcTs++ZHN3q4X6YQF5vTstagTTOCIQi8VSygctIiJSPAo3hXC6cHMldhukJTrCzpn9ELcGDq+CU7svbRtcA2pe7+jVqdkBKtcFiwXDMIhPzjwfeJLZeiyZ7ceSOXOZwNMwPIBBbWvQr0V1gnw8yuAARURErkzhphAuF24Kkn4a4lY7gk7cKojfAsZFTzD2q3o+6Fzv6OEJa5p355ZhGBxLOsf2v1zSWnfwDFm5jn14uVvpFR3B4LY1aFUzRL05IiJiKoWbQpSbcHOxrFQ4su582FkNR38H20Vjb7wCIardn5eyqrcEd6+8t5Mzcpi56ShT1x1h94nUvPX1Qv0Z1LYGd7SsTrCvZ1kdkYiISB6Fm0KU23BzsdwsOLbR0atzeBXErYXs1Pxt3LwgsvWfA5Sj2oJXAIZhsOlIElPXxvHj1uNk5jh6czzdrfRqFsGgtjVoU0u9OSIiUnYUbgpRYcLNxew2OLHdEXQuLBmn8rdx84I2D8KNT4FfFQBSMnOYvekYU9YdYWd8Sl7TOlX9zvfmRFLJT705IiJSuhRuClFhw83FDANO78sfdpLjHO95+kPsSMfiHXS+ucHWo8lMXRfHnC3Hych2PL/H081K96bhDGpbg/bXVVJvjoiIlAqFm0Io3BTAMGD/Ilj0L8fgZACfELjhSWj7EHj45DVNzcxhzpbjTF0Xx/Zjf/bm1K7ix8A2UfRvFUkVf6+LP0FEROSqKdwUQuHmCgwD/pgNi1+DU3sc6wIi4KZnoeVQcMt/e/i2o8lMXR/H7E3HSD/fm+PhZuHWJuEMbluD2OsqY7WqN0dERK6Nwk0hFG6KyJYLW7+DJa//ebkqpBbc8jw07X/JZKDpWbn8eL43Z8vR5Lz1NSv7MrBNDQa0iqRqgHpzRETk6ijcFELhpphys2DDl7DsTUhPdKwLbQyd/g8a9ITLjLHZcTyZb9cdYdamY6RmOZ6u7G610Kd5NV7q3UQPBxQRkWJTuCmEws1Vyk6HtZ/Cyvch83zPTPXW0PlFuK7jZTfJyM5l7tZ4pq6LY1NcEgA1KvnyyZCWNK0eVEaFi4hIeaBwUwiFm2t07iys+hDWjIOcDMe62h0dISeydYGbbTh8lsenbuJY0jk83a280qcJA9tE6e4qEREpEoWbQijclJDUE7D8bfh9IthzHOsa9HJcrgprfNlNkjKyefr7LSza5bi8dUeL6vz79qb4erqXVdUiIuKiFG4KoXBTwpLiYMkbsGXK+bmtLBB9F9z8D6h03SXN7XaD8csO8Ob8XdgNqB/mzydDWlE31L/saxcREZehcFMIhZtScnIPLP634zZyAKu749bxm/4OgRGXNF9z4DSPTd3EydQs/DzdGNs/mj4x1cq4aBERcRUKN4VQuCllxzfBb/+GfQsdr929HQ8BvOFJ8K2Ur2liaiZPTN3M6gOnARgaW5PnezXCy93t4r2KiEgFp3BTCIWbMnJopeNpx0fWOF57BjjmrLp+NFitec1ybXbeXbiHjxfvByAmMoiPBrckqpKvCUWLiIizUrgphMJNGTIM2LsAfvsXJGxzrKvfHe6YkDdn1QWLdyXy5PebScrIIcjHg3fuiqFzozATihYREWdUnO9va6HvilwLiwXq3woPLYM+HzkuUe35BT7vAqf25Wt6S8NQ5j52AzFRwSSfy+GBL3/njV92kWuzm1S8iIi4KoUbKX1WK7T8G9w3DwKrO+as+qwT7F2Yr1lkiC/THo5lWIdaAIxbsp8hn68lMSXThKJFRMRVKdxI2aneEoYvhqh2kJUM3wyAFe85Ll+d5+lu5eU+TfhocAv8PN1Ye/AMPT9Ywer9p82rW0REXIrCjZStgDC490fHbeIYsPAl+GE45JzL1+y26GrMeewGGoQFcCotiyGfr+Hjxfuw2yvUEDEREbkKCjdS9ty9oPcH0PMtx/Nwtk2Did0g6Ui+ZnWq+jNr5PUMaBWJ3YA35+/mgS/XczY926TCRUTEFSjciDksFmg7HIbOBt/KEL8FPrsFDq/O18zH04237ozhv/2j8XK3snj3SW77cAWbjySZU7eIiDg9hRsxV60b4KElENYM0k/Cl73h90mXNLurTRQzR1xPrcq+HEs6x52fruLLVYeoYE8yEBGRIlC4EfMF14AH5kOT2x2TcM4dDXOfhNz8l58aVwtkzmM30KNpODk2g5fm7GDU1E2kZeWaU7eIiDglhRtxDp5+MGASdH4RsDhmG/9fX0g7ma9ZoLcHnwxpyQu3NcbdauGnrfH0+WgFuxNSzalbREScjsKNOA+LBW58GgZ965iuIW4VTLjZMR4nXzMLD9xQm+8ejiUiyJsDJ9Pp+/EKZm8+Zk7dIiLiVBRuxPk06A7Df4PKdSHlKHzRDbZNv6RZq5oh/PT4jdxUvyqZOXae+HYzHy/ep3E4IiIVnMKNOKeq9eHBRVC3K+SegxkPwMKXwW7L16ySnyeTh7XhoZuuAxy3i78wezs2PQ9HRKTCUrgR5+UTDIO/c8wkDrDiXZhyN5xLytfMarXwz56NeKl3YywW+HpNHI98vYFz2baL9ygiIhWAwo04N6sbdH0F+n/hmHhz3wL4vDOc3HNJ0/uur80ng1vi6W5lwR8nGPz5Gs7ogX8iIhWOwo24hmYD4P5fIDASTu9zBJw98y9p1qNZBN882I4gHw82xSUxYNwq4k5nmFCwiIiYxdRwM27cOKKjowkMDCQwMJDY2FjmzZtXYPvJkydjsVjyLd7e3mVYsZiqWgvHA/9qxEJWiuMS1fK38028CdCmViVmPBpL9WAfDpxK545xK9l2NNmcmkVEpMyZGm4iIyN5/fXX2bBhA7///judOnWib9++7Nixo8BtAgMDiY+Pz1sOHz5chhWL6fyrwtA50Oo+wIBF/4Lp90N2er5mdUMD+GFEBxpFBHIqLZu7J6xmye5Ec2oWEZEyZWq46d27Nz179qRevXrUr1+f1157DX9/f9asWVPgNhaLhfDw8LwlLCysDCsWp+DuCb3fg17vOCbe3PHD+Yk34/I1Cwv05vuH23ND3SpkZNt44Mvf+f73I5ffp4iIlBtOM+bGZrPx7bffkp6eTmxsbIHt0tLSqFmzJlFRUVfs5QHIysoiJSUl3yLlRJsH4N4fwbcKJGyD8R1h/+J8TQK8PZg4rA13tKiOzW7w9+lb+WDRXj0LR0SkHDM93Gzbtg1/f3+8vLx45JFHmDlzJo0bN75s2wYNGjBx4kRmz57N119/jd1up0OHDhw9erTA/Y8dO5agoKC8JSoqqrQORcxQs4NjHE5EDJw7A1/fAcvfyTcOx9Pdytt3xTDi5joAvLNgD/+cuY1cm92kokVEpDRZDJP/CZudnU1cXBzJyclMnz6dzz//nKVLlxYYcP4qJyeHRo0aMWjQIF599dXLtsnKyiIrKyvvdUpKClFRUSQnJxMYGFhixyEmy8mEn5+GTV87Xje8Dfp9At5B+Zp9tfoQL83Zgd2ATg1D+WhwC3w93U0oWEREiiMlJYWgoKAifX+bHm4u1qVLF+rUqcP48eOL1P7OO+/E3d2dqVOnFql9cX444mIMAzZ+CT8/C7Zsx/QNd38NoY3yNft1RwKPTd1EVq6dmMggvhjWhir+XiYVLSIiRVGc72/TL0tdzG635+tpKYzNZmPbtm1ERESUclXiEiwWaDUM7vvL83A+6wzbZ+RrdmuTcKYMb0+IrwdbjibTf9wqDp1Kv/w+RUTE5ZgabsaMGcOyZcs4dOgQ27ZtY8yYMSxZsoQhQ4YAMHToUMaMGZPX/l//+he//vorBw4cYOPGjdxzzz0cPnyYBx980KxDEGcU2QoeXgq1O0JOuuNW8V/+CbacvCataoYw/dEORFXy4fDpDPqPW8XmI0nm1SwiIiXG1HCTmJjI0KFDadCgAZ07d2b9+vXMnz+frl27AhAXF0d8fHxe+7NnzzJ8+HAaNWpEz549SUlJYdWqVUUanyMVjF8VuOeHP+elWvMx/K8vpP35rJs6Vf2Z8WgHmlUP4nR6NoMmrGHRzhPm1CsiIiXG6cbclDaNuamA/pgDs0ZAdioERMCdX0KNdnlvp2flMuKbjSzdcxKrBV67vRmD2tYwsWAREbmYS4+5ESlxjfvA8N+gSgNIjYfJvWDdZ3m3i/t5ufP5va25s1UkdgPG/LCNd37drWfhiIi4KIUbqRiq1ncEnMb9wJ4DPz8DMx+GbMekmh5uVv47IJrHO9cD4IPf9vH36VvJ0bNwRERcjsKNVBxe/nDnZLj132Bxg63fwRe3wpkDgGNqj6e61mfsHc2wWmDahqM8+OXvpGflmlu3iIgUi8KNVCwWC3R4DIbOBr+qcGIbTLgZ9szPazKobQ0+G9oaHw83lu45yd0TVpOYmmlezSIiUiwKN1Ix1b4RHloKkW0gMxmm3AWLx4LdcRmqc6Mwpj7Unkp+nmw/lkL/cas4cDLN5KJFRKQoFG6k4gqqDsN+gjbnn5O09HVHyMk4A0DzqGB+eLQDNSv7cuTMOQZ8upotehaOiIjTU7iRis3dC3q9Df0+BXdv2LfAcZkqfisAtar45T0L50x6NoM+W8PyvSfNrVlERAqlcCMC0HwQPLAAgmtC0mH4oitsdsxXVsXfi6kPteeGulXIyLZx/+T1zNly3OSCRUSkIAo3IhdERMNDS6BuV8jNhFmPwE9PQ242/l7ufDGsNbdFR5BjM3ji201MXnnQ7IpFROQyFG5E/sq3Egz+Hjr+w/F6/ecwuSekHMfL3Y0PBrbg3tiaGAa8/OMfvK2H/YmIOB2FG5GLWa1wyxhHyPEOgqPrYfxNsPErrPZsXu7ThKe61gfgw9/28c+Z27HZFXBERJyFwo1IQep3c1ymCmsK6Sdhzih4LxrLyvd5/PpQXru9KVYLTF0Xx4hvNpCZYzO7YhERQRNnml2OuILsDMflqTXjIPX8QGLPAGg9jMUhA3h4VjzZNjvtalfis3tbE+jtYW69IiLlUHG+vxVuRIoqNxu2T4eVH8DJnY51Vg8Sa/fhoX0d2JwVQaOIQL68vw2hAd7m1ioiUs4o3BRC4Uaumd3ueB7Oyg/g8Iq81cssrfgwsxcnglvw1YPtqFnZz8QiRUTKF4WbQijcSIk6+jusfB92/gg4fpU22usy1b0f9943kqZRlcytT0SknFC4KYTCjZSK0/th1YcYm6dgsWUBcMiIwNZ+JHW6DAcPXaYSEbkWCjeFULiRUpWWSNaqceSsnoC/4ZhoM8urCl7XPwptHgCfEJMLFBFxTQo3hVC4kbKQmZ7MnIlv0OHUd0RaTjlWevhBq2HQ/lEIjjK1PhERV6NwUwiFGykrNrvBS7M2k/r79zziPpdG1jjHG1Z3aNofOjwO4U3NLVJExEUU5/tbD/ETKSVuVguv3t6cmrfcR4/ssQzNfo79/q3Angtbv4NPr4ev+8PBZVCx/o0hIlKq1HMjUga+Wn2IF+fswDDg0XopPBPwC24754BhdzSo3gp6vQPVmptap4iIs1LPjYiT+VtsLT4c1AIPNwvj9gYyNOVR0h9eD20eBHdvOLYBPusEi16F3CyzyxURcWkKNyJl5Lboaky+ry1+nm6s3Heau6fFc6rjf2D0NmhyOxg2WP4WjO/oCDsiInJVFG5EytD1davw7UOxVPbzZPuxFAaMW8WRbH+4czLc9T/wq+qY2uHzLrDgJcjJNLtkERGXo3AjUsaaRQYx/dEORIb4cOh0BneMW8XO+BRo3BdGrIVmdzrG4qx8D8bfCEfWmV2yiIhLUbgRMUHtKn7MeLQDDcMDOJmaxV2frmb6hqMYvpWg/+cwcAr4h8GpPfDFrTD/ecg5Z3bZIiIuQeFGxCRhgd5893AsbWtXIjUrl2embeH+yeuJTz4HDXvBiDUQMwgwYPVHMO56OLza7LJFRJyewo2IiYJ8PJjyYDue7dYATzcri3ef5NZ3lvHtujgMnxC4/VMY/D0EVIMz+2FSD5j3D8hON7t0ERGnpefciDiJvSdSeXb6VjYfSQLgxnpVGHtHMyJDfOFcEvz6PGz62tE4pBb0+Qhq32hWuSIiZUrTLxRC4Uacmc1uMHHFQd76dTdZuXb8PN34R89GDGlbA6vVAvsWwpwnIOWoY4M2D0KXV8DL39zCRURKmcJNIRRuxBUcOJnG36dv5ffDZwGIva4yb/SPpkZlX8hMgQUvwIbJjsZBNaDvh3DdzabVKyJS2hRuCqFwI67CZjf4ctUh/jt/F5k5dnw83Ph79wbcG1vL0YtzYAnMeQySzk/I2WoYdH0VvPXftYiUPwo3hVC4EVdz+HQ6f5++lbUHzwDQplYI/x0QQ+0qfpCVBgtfhvWfORoHRkKf96FuF/MKFhEpBQo3hVC4EVdktxt8s/YwY+ftIiPbhpe7lWdubcD9N9TGzWqBg8thzig4e8ixQYt74NbXwCfYzLJFREqMwk0hFG7ElR05k8GYH7axYt8pAFrUCObNAdHUDQ1w3B6+6FVY+ylgOG4f7/0e1O9mas0iIiVB4aYQCjfi6gzD4Lv1R/j3TztJy8rF093K6C71eOjG63B3szoe9Dd7pOO5OADRA6HrvyAgzNzCRUSugcJNIRRupLw4nnSOMT9sY+mekwBERwbx5oAYGoQHQHYGLH4NVn8MnP8V9wqCkBoQXNOxhJz/M7iG4++efuYdjIjIFbhMuBk3bhzjxo3j0KFDADRp0oQXX3yRHj16FLjNtGnTeOGFFzh06BD16tXjjTfeoGfPnkX+TIUbKU8Mw2D6hqO8OvcPUjJz8XCz8Finejx6cx083KxwZD389CQkbLvyznyrXBp4gms6HhgYFAnuXqV+PCIiBXGZcPPjjz/i5uZGvXr1MAyDL7/8kjfffJNNmzbRpEmTS9qvWrWKm266ibFjx3LbbbcxZcoU3njjDTZu3EjTpk2L9JkKN1IenUjJ5PmZ21i4MxGAxhGBvHlnNE2qBTkaZGc4bhlPOgxnDzv+/OvfM5Ov8AkWCIj4S+A5H4Au/D0wEqyazUVESo/LhJvLqVSpEm+++SYPPPDAJe/dfffdpKenM3fu3Lx17du3p3nz5nz66adF2r/CjZRXhmEwZ8txXpqzg6SMHNytFkbcXIdRnerh6X6F4HEu6dLwc/bwn+tyMgrf3jMAwptCeDSEN4OIaKjaUL09IlJiivP97V5GNV2RzWZj2rRppKenExsbe9k2q1ev5qmnnsq3rlu3bsyaNavA/WZlZZGVlZX3OiUlpUTqFXE2FouFvs2rE1unMi/O2sEvOxL44Ld9zN9xgjfvjCY6MrjgjX2CHUtE9KXvGQakn7q0tyevB+gIZKdC3GrHcoHVwxFwIs4HnvBoRwDyDirhIxcRyc/0cLNt2zZiY2PJzMzE39+fmTNn0rhx48u2TUhIICws/x0fYWFhJCQkFLj/sWPH8sorr5RozSLOLDTAm3H3tOSnbfG8OHsHu0+kcvsnqxh1S11GdarrGItTHBYL+Fd1LJGtL33flgun9jjG9SRshfgtjj8zk+HENsfyVyG1HEEnIvp84ImGgHDH54iIlADTL0tlZ2cTFxdHcnIy06dP5/PPP2fp0qWXDTienp58+eWXDBo0KG/dJ598wiuvvMKJEycuu//L9dxERUXpspRUCKfTsnhxzg5+2hoPQLPqQbx7d4zjuTilyTAg+QjEb/1L6Nn654SfF/Or+mfvzoXQU6mOxvGISB6Xuizl6elJ3bp1AWjVqhXr16/n/fffZ/z48Ze0DQ8PvyTEnDhxgvDw8AL37+XlhZeXrvtLxVTZ34uPB7ekW5Pj/N/MbWw7lkyvD1bw9+4Nua/D+TmqSoPFcn7AcQ1odNuf6zPO/Bl0LoSeU3sg/STs/82xXODhB2FNoHpLaD/CMXBZRKQITO+5uVinTp2oUaMGkydPvuS9u+++m4yMDH788ce8dR06dCA6OloDikWuICE5k7/P2Mqy88/FaX9dJd66M4bIEF9zC8vOgMSdkLDlz9BzYgfknvuzjYcvdPo/aPswuJn+bzIRMYHL3C01ZswYevToQY0aNUhNTc27tXv+/Pl07dqVoUOHUr16dcaOHQs4bgXv2LEjr7/+Or169eLbb7/lP//5j24FFykiwzD4Zm0cr/20k3M5Nvy93Hmpd2MGtIrE4kxjXmy5cHqfI+hsmASHVzrWRzSHPh9ARIyp5YlI2SvO9/dVXdA+cuQIR4/+ee183bp1jB49mgkTJhRrP4mJiQwdOpQGDRrQuXNn1q9fnxdsAOLi4oiPj89r36FDB6ZMmcKECROIiYlh+vTpzJo1q8jBRqSis1gs3NO+JvOeuJGWNYJJy8rl2elbeeirDZxKy7ryDsqKmzuENoToO+HeudD7A8ddVvGbYcIt8OsLjh4fEZHLuKqemxtvvJGHHnqIv/3tbyQkJNCgQQOaNGnC3r17eeyxx3jxxRdLo9YSoZ4bEQeb3WD8sv28u2APOTaDyn6e/OeOZnRrUvAYNlOlnoBfnoMdMx2vg2s6Jgat08nUskSkbJR6z8327dtp27YtAN9//z1NmzZl1apVfPPNN5cdKyMizsfNamHEzXWZPfIGGoYHcDo9m4e/2sDT328hJTPH7PIuFRAGd06GQd86noicdBi+uh1+eBjST5tdnYg4kasKNzk5OXl3IC1cuJA+ffoA0LBhw3yXkUTE+TWuFsjsUdfzSMc6WCwwY+NRery3nFX7T5ld2uU16AEj10C7RwALbP0WPmoNW7513IIuIhXeVYWbJk2a8Omnn7J8+XIWLFhA9+7dATh+/DiVK1cu0QJFpPR5ubvxjx4N+f7hWGpU8uVY0jkGf7aWV37cQWaOzezyLuUVAD3egAcXQWgTOHcGZj7s6Mk5c9Ds6kTEZFcVbt544w3Gjx/PzTffzKBBg4iJcdy5MGfOnLzLVSLietrUqsS8J25kcLsaAExaeYheHyxny5EkcwsrSGQreHgpdH4R3LzgwGL4JBZWvu+440pEKqSrvhXcZrORkpJCSEhI3rpDhw7h6+tLaGhoiRVY0jSgWKRoFu9O5LnpW0lMzcLNarn66RvKyun9MHc0HFzmeB3eDPp8CNVamFqWiJSMUn/Ozblz5zAMA19fx8O/Dh8+zMyZM2nUqBHdunW7uqrLiMKNSNElZWTzf7O2M7esp2+4WoYBm6fA/H9CZhJYrI6nG9/yT/D0M7s6EbkGpR5ubr31Vu644w4eeeQRkpKSaNiwIR4eHpw6dYp33nmHRx999KqLL20KNyLFN2fLcV6YtZ3kczl4uVtLf/qGa5V2EuaPgW3THK+DasBt70K9LubWJSJXrdRvBd+4cSM33ngjANOnTycsLIzDhw/zv//9jw8++OBqdikiTqxPTDV+ffImOtavSlaunVfn/sHgz9dw9KyTPkjPvyr0/xyGTHcEm+Q4+KY/zHjQEXxEpFy7qnCTkZFBQICjW/rXX3/ljjvuwGq10r59ew4fPlyiBYqIcwgL9GbyfW34d7+m+Hi4sebAGbq/t5xpvx/Byaao+1O9ro7bxmNHOS5RbZsGH7eBTd/otnGRcuyqwk3dunWZNWsWR44cYf78+dx6662AYzoFXeoRKb/+On1Dq5ohedM3PDNtK7k2u9nlXZ6nH3R7DYb/BuHRcO4szB4B/+vjGIQsIuXOVY25mT59OoMHD8Zms9GpUycWLFgAwNixY1m2bBnz5s0r8UJLisbciJQMm91gwrIDvPXrbmx2g17REbx3d3PnvZsKHLeHr/kYFo91zDru7g1th0OV+uAfBv6hjj/9qoKbh9nVishflMms4AkJCcTHxxMTE4PV6vif2bp16wgMDKRhw4ZXs8syoXAjUrLm70hg1JSN5NgMujYO46PBLfBydzO7rMKdOQhzn3Q8F6cgPpXyBx7/0Px/9zv/d9/KYHXiQCdSTpRJuLngwuzgkZGR17KbMqNwI1LyFu9K5OGvN5Cda+fmBlX59J5WeHs4ecAxDMcknPsWQXoipJ2AtERIPwn2YjwA0OLm6Onxr3o++Pw1/IRCjVgIql56xyFSQZR6uLHb7fz73//m7bffJi0tDYCAgACefvppnn/++byeHGekcCNSOlbsPcWD/1tPZo6dG+pWYcLQVvh6uptdVvHZ7Y5xOWknHEv6yT//npb4l+UEZJwGrvC/UHdvuOkZ6PAEuHuWySGIlEelHm7GjBnDF198wSuvvML1118PwIoVK3j55ZcZPnw4r7322tVVXgYUbkRKz9oDp7l/8nrSs220rV2JicPa4O/lggGnqGw5kH7qfM9P4l9C0PlAdHovJGxztK3SwPGsnVrXm1uziIsq9XBTrVo1Pv3007zZwC+YPXs2I0aM4NixY8XdZZlRuBEpXRsOn2XYxHWkZuXSokYwk+9rS5BPBR2caxiwbbrjgYLp55+v0/we6Pov8NMkwyLFUeoP8Ttz5sxlBw03bNiQM2fOXM0uRaScaFUzhG+GtyPIx4NNcUnc8/lakjKyzS7LHBYLRN8Jo9ZDq/sc6zZ/DR+1hk1f61k7IqXkqsJNTEwMH3300SXrP/roI6Kjo6+5KBFxbdGRwUwZ3o5Kfp5sO5bMoM/Wcjoty+yyzOMTAr3fgwcWQGgTOHcGZo+Eyb3g5G6zqxMpd67qstTSpUvp1asXNWrUIDY2FoDVq1dz5MgRfv7557ypGZyRLkuJlJ09J1IZ/NlaTqVlUS/Un2+GtyM0wNvsssxly4E1n8CS1yEnA6wecP0TjkHHHj5mVyfitEr9slTHjh3Zs2cPt99+O0lJSSQlJXHHHXewY8cOvvrqq6sqWkTKn/phAXz/cHvCA73Zm5jGwPFriE8+Z3ZZ5nI7H2ZGroX63cGeA8vfgk9iHbeli8g1u+bn3PzVli1baNmyJTabraR2WeLUcyNS9uJOZzDoszUcSzpHVCUfpjzYnqhKvmaXZT7DgF1z4ee/Q+pxx7qm/aHbWAgIM7c2ESdT6j03IiLFUaOyL9893J6alX05cuYcAyes4dCpdLPLMp/FAo16w6h10H6EY3LP7TPgozaw7jOwO+8/FEWcmcKNiJSJyBBfvnsoluuq+nEs6Rx3T1jNvsQ0s8tyDl4B0H0sDF8M1VpAVjL8/Ax80RXit5pdnYjLUbgRkTITHuTNdw/F0iAsgBMpWQycsJrdCalml+U8qjWHBxdBjzfBMwCObYAJN8P85yFLQVCkqIo15uaOO+4o9P2kpCSWLl2qMTciUqgz6dnc8/la/ohPIcTXg68eaEfT6kFml+VcUuLhl3/AH7McrwMjoed/oWEvU8sSMUupPaH4vvvuK1K7SZMmFXWXZU7hRsQ5JGfkMHTiWrYcTSbQ253/PdCO5lHBZpflfPYugJ+ehqTDjtcNekGPNyA4yty6RMpYmc4K7moUbkScR2pmDvdNWs/vh8/i7+XO5Pva0LpWJbPLcj7ZGbDsTVj1gWPGcg8/uGUMtHsU3Mrx3F0if6G7pUTEJQR4e/Dl/W1pf10l0rJyGTpxHav3nza7LOfj6QtdXoJHVkCNWMhJh1//zzEeZ9t0yEw2u0IRp6KeGxEx3blsGw999TvL957Cy93KZ0Nbc1P9qmaX5Zzsdsf8VAtehHNnHeusHlD7Rsd4nAY9IbCauTWKlAJdliqEwo2Ic8rMsTHym40s2pWIp5uVcfe0pHMjPciuQOmnHNM47JwLpy6an6paS0fQaXgbVG3geJ6OiItTuCmEwo2I88rOtfP41E38siMBDzcLHw5qQfemEWaX5fxO7YVdP8Hun+HIOuAv/1uvVOfPoBPZGqxuppUpci0UbgqhcCPi3HJsdp76fgs/bjmOm9XCu3c3p0+MLrMUWeoJ2DPPEXYOLAFb9p/v+YVCgx6OsFO7I3hU8ElMxaUo3BRC4UbE+dnsBn+fvpUZG49itcBTXevzSMc6uLvpHohiyUqFfQsdQWfPr44nH1/g4Qf1ujh6dOp1BZ8Q8+oUKQKFm0Io3Ii4Brvd4P9mb2fK2jgAWtUM4Z27YqhZ2c/kylxUbjYcXuEIOrt+/nOiTgCrO9S6wRF0GvSAoEjz6hQpgMJNIRRuRFyHYRjM3HSMl2bvIDUrFz9PN17s3Zi7Wkdh0SDZq2e3Q/ymP4POyZ35349o/mfQCWuiAcniFBRuCqFwI+J6jp7N4Knvt7Du4BkAujQK4/X+zaji72VyZeXE6f3ng85PcGQt+QYk+1Z2PFunRizU7ADh0XpwoJhC4aYQCjcirslmN/hixQHemr+HbJudyn6evN4/mq6Ndbt4iUpLhN3nByQfXAq5mfnf9/SHqLZQo4Mj7FRvpYHJUiYUbgqhcCPi2nbGp/Dkd5vZdX428YFtovi/2xrj76XehBKXmw3xm+HwSji8GuLW5B+UDODm6Qg4F3p2otqCtyZBlZLnMuFm7Nix/PDDD+zatQsfHx86dOjAG2+8QYMGDQrcZvLkyZdM4Onl5UVmZmYBW+SncCPi+jJzbLyzYA+fLT+AYUCNSr68e3cMrWpqXqpSZbdB4h+OoHN4JcSthrQT+dtYrBDWFGpeDzVjHT08/nratFw7lwk33bt3Z+DAgbRp04bc3Fz++c9/sn37dv744w/8/C5/R8TkyZN54okn2L37zydyWiwWwsKK1jWtcCNSfqzef5pnpm3hWNI5rBYYcXNdHu9cD0933TJeJgwDzhyAw6scS9wqOHvo0naV6zl6dS4swTXKvFRxfS4Tbi528uRJQkNDWbp0KTfddNNl20yePJnRo0eTlJR0VZ+hcCNSvqRk5vDy7B38sOkYAE2rB/Le3c2pGxpgcmUVVMrx80FntePPxD8ubRMY6ejVqdrQ0dNjsTruyLJYActlXp9fLvteIa+rt4SQWmV59FKKivP97VQXqZOTHddyK1UqvGs5LS2NmjVrYrfbadmyJf/5z39o0qTJZdtmZWWRlZWV9zolJaXkChYR0wV6e/DO3c3p3CiM52dtY/uxFHp9sIJ/9GjIvbG1sFp1G3OZCqwGzQY4FoCMM46xOnGrHJez4jdDylHYNq30a7FYoVFv6PC4Y+oJqTCcpufGbrfTp08fkpKSWLFiRYHtVq9ezd69e4mOjiY5OZm33nqLZcuWsWPHDiIjL33w1Msvv8wrr7xyyXr13IiUPydSMnl2+laW7TkJwI31qvDmgBjCg3Q3j9PIToej6x29OinHzt91boBhd1zmMuxXeG0UrX1WGhzf+Ofn1ugA1z8O9bqBVZctXZFLXpZ69NFHmTdvHitWrLhsSClITk4OjRo1YtCgQbz66quXvH+5npuoqCiFG5FyyjAMvl5zmNd+3klmjp0gHw/+3a8pvTU/VcVz4g9Y/RFs/R7sOY51VepD7CiIvlu3sLsYlws3o0aNYvbs2SxbtozatWsXe/s777wTd3d3pk6desW2GnMjUjHsP5nGU99tZstRx+Xuvs2r8a8+TQny9TC5MilzKcdh7Xj4fdKft7L7hUK7h6D1A+Cru+xcQXG+v03tmzMMg1GjRjFz5kx+++23qwo2NpuNbdu2ERERUQoVioirqlPVn+mPduDxzvVws1qYvfk43d9fxsp9p8wuTcpaYDXo+go8uR1ufc0xoDk9EX77N7zbBH7+++Xv8hKXZWrPzYgRI5gyZQqzZ8/O92yboKAgfHx8ABg6dCjVq1dn7NixAPzrX/+iffv21K1bl6SkJN58801mzZrFhg0baNy48RU/Uz03IhXPxrizPPXdZg6dzgDggRtq82y3Bnh7uJlcmZjClgM7ZsLKD+DENsc6ixUa93UMPq7e0tz65LJcpudm3LhxJCcnc/PNNxMREZG3fPfdd3lt4uLiiI+Pz3t99uxZhg8fTqNGjejZsycpKSmsWrWqSMFGRCqmljVC+PmJGxnSzvF8lS9WHKT3hyvYcTz5CltKueTmAdF3wSPL4W+zoE5nx0DkHTPhs1tgUi/YM98xwai4JKcYc1OW1HMjUrH9tusEf5++jVNpWXi4WXiya30evqkObrplvGJL2A6rPoTt08Ge61hXpQF0eMwRhNw1SavZXG5AcVlSuBGR02lZ/HPmNubvcEwd0KFOZd69uzlhgbp7psJLPgZrx8HvkyHbMX8Z/mHQ7mFofT/4hJhaXkWmcFMIhRsRAccNDdM2HOWl2Ts4l2Ojkp8nb98Vwy0NQs0uTZxBZjJs+BLWjIPU4451Hn7QcijEjtAUEiZQuCmEwo2I/NW+xDQem7qJnfGOp5c/eENt/t69oeanEofcbNjxg+OS1YntjnUWN2jSD9qPhMhWppZXkSjcFELhRkQulplj4/V5u5i86hAA0ZFBfDCwBbWqXH4CX6mADAP2/warPoADS/5cX70VtH3YEXY0LqdUKdwUQuFGRAry644E/j5jK0kZOfh5uvGfO5rRt3l1s8sSZxO/FdZ8AttngC3bsc6vKrS6zzEuJ1DPXSsNCjeFULgRkcLEJ5/jiambWXfoDAADWkXySp8m+Hk51TzD4gzSTsLGybB+4p/jcqzujufltH0Yotqen81cSoLCTSEUbkTkSnJtdj78bR8f/rYXuwHXVfXjw0EtaFItyOzSxBnZcmDnj7BuAsSt/nN9RIwj5DTtr3msSoDCTSEUbkSkqNYcOM3obzeTkJKJp5uV53s1YmhsTSz617gUJH6LI+RsnQa285M2+1aGVsMc81gF6TLn1VK4KYTCjYgUx9n0bJ6dvoWFOxMB6No4jP/2jybEz9PkysSppZ+GjV/C+i8g5ahjncUNGt0G7R6BGrG6ZFVMCjeFULgRkeIyDIPJqw4x9uddZNvsRAR5897dzWl3XWWzSxNnZ8uF3T87enMOLf9zfVgzx6zkze4EDx/z6nMhCjeFULgRkau1/Vgyj0/dxIFT6Vgt8HjnejzWqZ6mbpCiObHDEXK2fAe55xzrfEIcDwZs86AeDHgFCjeFULgRkWuRnpXLi7N3MGOj41JDu9qVeG9gcyKC9K9vKaKMM7Dpa1j/GSTFOdZZrNCgp2Oah1o36pLVZSjcFELhRkRKwsxNR/m/mdtJz7YR7OvBWwNi6NI4zOyyxJXYbY7Zx9d+CgeX/rk+tLFjXE6Lv4FVT8q+QOGmEAo3IlJSDp5K5/Gpm9h2LBmAYR1qMaZnQ7zc3UyuTFxO4q7zl6y+hZx0x7rou6HvJ+CmZyyBwk2hFG5EpCRl59r57y+7+HzFQQAaRwTy4eAW1Knqb3Jl4pLOJcGGSbDoVTBs0KAXDJio5+RQvO9v9XeJiFwDT3cr/3dbYyYNa0MlP0/+iE+h94crmLHhqNmliSvyCYYbnoSB34CbF+z+CabcCVmpZlfmUhRuRERKwC0NQ5n3xI3EXleZjGwbT0/bwpPfbSYtK9fs0sQVNegB98wAT384uAz+188xEFmKROFGRKSEhAV68/WD7Xjm1vq4WS3M3HSMbu8u45ft8VSwEQBSEmrfCPfOcdwufux3mNwLUhPMrsolKNyIiJQgN6uFUZ3q8d1D7ake7MOxpHM88vVGhk5cx77ENLPLE1dTvRXcNw/8wyHxD5jYHc4eNrsqp6dwIyJSClrXqsTCpzryeKe6eLpbWb73FN3fW8Z/ft5JamaO2eWJKwltBPf/AiG14OxBmNjNcXeVFEh3S4mIlLK40xn8a+4fLNx5AoCqAV78s2dD+jWvrkk4pehS4uGr2+HkTvCp5BiTU72l2VWVGd0KXgiFGxExy+Jdibzy4w4Onc4AoE2tEF7u04Qm1YJMrkxcRsYZ+Lo/HN8IngEw+FuodYPZVZUJhZtCKNyIiJmycm18vvwgH/22j3M5NqwWGNKuJk/fWp9gX800LkWQlQpTBzkm4nT3hrv+B/W7mV1VqVO4KYTCjYg4g+NJ5/jPzzuZuzUegBBfD57t1pC720RpIk65spxMmDYM9swDqzvcPh6aDTC7qlKlcFMIhRsRcSar9p/ilTl/sPuE4yFtzaoH8UrfJrSsEWJyZeL0bDkwawRs+x6wwG3vQOv7za6q1CjcFELhRkScTa7NzldrDvPOr3tIPf/QvwGtInmue0OqBniZXJ04Nbsd5j0L6z93vO7ysuMJx+WQwk0hFG5ExFmdTM3iv7/sYtr5qRsCvNwZ3bU+Q2Nr4uGmJ3dIAQwDfnsVlr/teH3Dk9D5JShnd+Ip3BRC4UZEnN3GuLO8NHtH3mzj9cP8eblPEzrUqWJyZeLUVrwHC19y/L31/dDzbbCWn1CscFMIhRsRcQU2u8H3vx/hv7/s4myG46F/vaIjeL5nI6oF+5hcnTit3yfB3CcBA5rdCf3GgZuH2VWVCIWbQijciIgrScrI5p0Fe/h6zWHsBvh4uDGqU10evLE2Xu5uZpcnzmj7DPjhIbDnQv0ecOck8HD9QKxwUwiFGxFxRTuOJ/PynB2sP3QWgFqVfXmxd2M6NQwzuTJxSnt+he//BrmZUOtGGDQVvALMruqaKNwUQuFGRFyVYRjM3nyc//y8k8TULADuaFmd/9zeDG8P9eLIRQ6thCl3Q3YqVGvpmK7Bt5LZVV214nx/l5+RRiIi5ZzFYqFfi+r89szNPHzTdbhZLfyw8Rh3j19NQnKm2eWJs6l1PQz7EXwrO6ZrmNQDUo6bXVWZULgREXEx/l7ujOnZiK/ub0uwrwdbjibT+6MVbDh8xuzSxNlUawH3zYOAanByF0zsDmcOml1VqVO4ERFxUR3qVmHOyBtoGB7AydQsBk5Yw3fr48wuS5xN1QZw/y8QUhuSDjsCzok/zK6qVCnciIi4sBqVfZnxaAd6NA0nx2bw3IxtvDR7Ozk2u9mliTMJqekIOKFNIC0BPr0BvroDNn0DmclmV1fiNKBYRKQcMAyDj37bx9sL9gDQ/rpKfDy4JZX9NX2D/EXGGZh+PxxY/Oc6Ny+o19Ux8Wb97k5727juliqEwo2IlGcL/jjBk99tJi0rl+rBPkwY2oom1YLMLkuczal9sOMH2DYdTu3+c72nPzTo6Qg6dTo51QMAXeZuqbFjx9KmTRsCAgIIDQ2lX79+7N69+4rbTZs2jYYNG+Lt7U2zZs34+eefy6BaERHn17VxGDNHdKBWZV+OJZ1jwLjV/LQ13uyyxNlUqQsd/w4j18IjK+D60RBUA7LTHLOMT7kL3qoHPz4BB5eD3WZ2xcVias9N9+7dGThwIG3atCE3N5d//vOfbN++nT/++AM/P7/LbrNq1Spuuukmxo4dy2233caUKVN444032LhxI02bNr3iZ6rnRkQqguSMHEZN3cjyvacAGHlLHZ7u2gCrtXxNpiglyDDg6HpHb86OmZCe+Od7ARHQ5A5o2h+qtzRlUk6XvSx18uRJQkNDWbp0KTfddNNl29x9992kp6czd+7cvHXt27enefPmfPrpp1f8DIUbEakobHaD//6yi/HLDgDQuWEo7w5sTqC381xqECdly4VDy2H7dPjjR8j6y6DjkNqOkNNsAIQ2KrOSXOay1MWSkx0/vEqVCn6C4urVq+nSpUu+dd26dWP16tWXbZ+VlUVKSkq+RUSkInCzWhjTsxHv3d0cL3cri3Yl0u/jlRw4mWZ2aeLs3Nyhzi3Q92N4di8MnOoINB6+cPYgLH8LPmkPn3SAZW853bNznCbc2O12Ro8ezfXXX1/o5aWEhATCwvLPpRIWFkZCQsJl248dO5agoKC8JSoqqkTrFhFxdv1aVGf6Ix2ICPLmwMl0+n68ksW7E6+8oQiAuxc07AkDJsKz+6D/F44JOa0ekLgDfnsVPmgOn3WGNeMg9fLfx2XJacLNyJEj2b59O99++22J7nfMmDEkJyfnLUeOHCnR/YuIuIJmkUHMGXUDrWuGkJqZy/2T1/Pp0v040cgEcQWefo7LUYO/hWf2QJ8PoXZHsFjh2O/wyz/gnUbwdX+wm/esJacIN6NGjWLu3LksXryYyMjIQtuGh4dz4sSJfOtOnDhBeHj4Zdt7eXkRGBiYbxERqYiqBngxZXh7BrWtgWHA6/N28cS3mzmX7Vp3woiT8K0ELYfCvXPgqV3Q/Q2IbAOG3dGrYzUvYpgabgzDYNSoUcycOZPffvuN2rVrX3Gb2NhYFi1alG/dggULiI2NLa0yRUTKDU93K2PvaMa/+zXF3WphzpbjDPh0FceSzpldmriygDBo/wg8uBCe2AJdXjK1HFPDzciRI/n666+ZMmUKAQEBJCQkkJCQwLlzf/6SDR06lDFjxuS9fuKJJ/jll194++232bVrFy+//DK///47o0aNMuMQRERc0j3ta/LNg+2o7OfJjuMp9PlwBesOauJNKQEhtcr0LqrLMTXcjBs3juTkZG6++WYiIiLylu+++y6vTVxcHPHxfz6AqkOHDkyZMoUJEyYQExPD9OnTmTVrVpGecSMiIn9qd11lZo+6nsYRgZxOz2bwZ2v4es1hs8sSuWZO9ZybsqDn3IiI5Hcu28az07cw9/yTjAe3q8HLvZvg6e4UwzJFABd+zo2IiJQ9H083PhzUgue6N8RigSlr4xjy+RpOpmaZXZrIVVG4ERERLBYLj95ch4n3tiHA2531h85y24fLeWv+bjYcPoPNXqE6+cXF6bKUiIjks/9kGsP/9zsHTqbnrQvy8eDGelW4uUEoHetXpWqAl4kVSkXksnNLlQWFGxGRK8vIzuXnbQks2Z3I8r2nSD6Xk+/9ZtWDuLlBVW5uUJXmUSG4aUJOKWUKN4VQuBERKZ5cm53NR5JYsvskS/Yksv1Y/jn6gn09uLFeVW6uX5WODapSxV+9OlLyFG4KoXAjInJtElMzWbr7JEv2nGT5npOkZObmez86Muh80AmleVSwenWkRCjcFELhRkSk5Fzo1Vm8O5Elu0+y4/ilvTo31XNcvrqpvnp15Oop3BRC4UZEpPQkpmSyZM9Jlu4+ybK9J0n9S6+OxQLR1YPo2CCUmxtUJSZSvTpSdAo3hVC4EREpG7k2Oxvjklhyvlfnj/j8vTqV/Tzp2jiMbk3Dub5OFT00UAqlcFMIhRsREXOcSLkwVsdxB9Zfe3UCvNzp3CiU7k3Dual+VXw93U2sVJyRwk0hFG5ERMyXY7Oz9sAZftkRz/wdJ/I9Ddnbw8rN9R1B55aGoQT5eJhYqTgLhZtCKNyIiDgXu91g05GzzNuWwC87Ejh69lzeex5uFjrUqUL3puF0bRymAckVmMJNIRRuREScl2EY7DiewvwdCczbnsC+xLS896wWaFOrEt2bhtOtSTjVgn1MrFTKmsJNIRRuRERcx77EVObvOMEv2xPYdiw533sxUcF0bxJO96bh1K7iZ1KFUlYUbgqhcCMi4pqOnMlg/o4E5u9I4PfDZ/nrt1eDsAC6N3UEnYbhAVgsusW8vFG4KYTCjYiI60tMzWTBH44endX7T5P7l1nLa1b2pXuTcAa1rUEt9eiUGwo3hVC4EREpX5Iyslm0M5FfdiSwbM9JsnLtAPh5uvHh4BZ0ahhmcoVSEhRuCqFwIyJSfqVn5bJk90kmrjzIhsNnsVrg+V6Nuf/6WrpU5eKK8/2tx0GKiEi54eflTq/oCKYOb8/draOwG/Dq3D94ftZ2cmx2s8uTMqJwIyIi5Y6nu5XX+zfj+Z6NsFhgyto4hk1aR3JGjtmlSRlQuBERkXLJYrEw/KbrmPC31vh6urFy32luH7eSQ6fSzS5NSpnCjYiIlGtdG4cx/ZEOVAvy5sDJdPp9spI1B06bXZaUIoUbEREp9xpXC2TWyOuJiQwiKSOHv32xlu/XHzG7LCklCjciIlIhhAZ6893DsfSKjiDHZvD3GVsZ+/NO7PYKddNwhaBwIyIiFYa3hxsfDmzB453rATB+2QEe/noD6Vm5JlcmJUnhRkREKhSr1cJTXevz3t3N8XS3suCPE9z56Wrik89deWNxCQo3IiJSIfVrUZ2pw9tTxd+TP+JT6PvRSrYcSTK7LCkBCjciIlJhtaoZwswR19MgLIDE1CzuGr+an7fFm12WXCOFGxERqdCiKvky/dFYbm5QlaxcOyO+2chHv+2lgs1OVK4o3IiISIUX4O3B50Nbc9/1tQB469c9PPX9FrJybeYWJldF4UZERARwd7PyUu8m/LtfU9ysFmZuOsbgz9ZyKi3L7NKkmBRuRERE/uKe9jX58r62BHi7s+HwWfp9vJI9J1LNLkuKQeFGRETkIjfUq8LMEddTs7IvR8+eo/8nq1iyO9HssqSIFG5EREQuo26oP7NGXE/b2pVIzcrl/snr+XLVIbPLkiJQuBERESlAiJ8nXz/QjjtbRWI34KU5O3hh1nZybXazS5NCKNyIiIgUwtPdyn8HRPOPHg2xWOCrNYe5b/J6ks/lmF2aFEDhRkRE5AosFguPdKzDp/e0wsfDjeV7T9Hrg+VsijtrdmlyGQo3IiIiRdStSTjTHoklqpIPR8+e485PVzN+6X7NLO5kTA03y5Yto3fv3lSrVg2LxcKsWbMKbb9kyRIsFsslS0JCQtkULCIiFV7T6kH89PiN9IqOINduMHbeLu6bvF7Pw3Eipoab9PR0YmJi+Pjjj4u13e7du4mPj89bQkNDS6lCERGRSwV6e/DRoBaMvaMZXu5Wlu45Sc/3l7Nq3ymzSxPA3cwP79GjBz169Cj2dqGhoQQHB5d8QSIiIkVksVgY1LYGLWuEMGrKRvYmpjHki7WMuqUuT3Suh7ubRn6YxSV/8s2bNyciIoKuXbuycuXKQttmZWWRkpKSbxERESkpDcIDmDPqBga2icIw4MPf9jHoszUcTzpndmkVlkuFm4iICD799FNmzJjBjBkziIqK4uabb2bjxo0FbjN27FiCgoLylqioqDKsWEREKgIfTzde7x/NB4Na4O/lzvpDZ+n5wXIW/HHC7NIqJIvhJHO6WywWZs6cSb9+/Yq1XceOHalRowZfffXVZd/PysoiK+vPQV4pKSlERUWRnJxMYGDgtZQsIiJyicOn03ls6ia2Hk0GYFiHWozp2RAvdzeTK3NtKSkpBAUFFen726V6bi6nbdu27Nu3r8D3vby8CAwMzLeIiIiUlpqV/Zj+SAcevKE2AJNXHeKOT1Zx8FS6yZVVHC4fbjZv3kxERITZZYiIiOTxdLfyf7c1ZuKw1oT4erDjeAq3fbCcWZuOmV1ahWDq3VJpaWn5el0OHjzI5s2bqVSpEjVq1GDMmDEcO3aM//3vfwC899571K5dmyZNmpCZmcnnn3/Ob7/9xq+//mrWIYiIiBSoU8Mw5j1xE098u4m1B88w+rvNrNh3in/1bYKvp6lfweWaqT03v//+Oy1atKBFixYAPPXUU7Ro0YIXX3wRgPj4eOLi4vLaZ2dn8/TTT9OsWTM6duzIli1bWLhwIZ07dzalfhERkSsJD/JmyvD2jO5SD6sFpm84Su8PV7AzXnfvlhanGVBcVoozIElERKQkrTlwmie+3cSJlCw83a28cFtj7mlXA4vFYnZpTq9CDSgWERFxFe2vq8y8J26iU8NQsnPtvDBrOyO+2agZxkuYwo2IiEgZquTnyRf3tub/ejXCw83CvO0J9Hx/ORs1w3iJUbgREREpYxaLhQdvvI4Zj3agRiVfjiU5Zhgft0QzjJcEhRsRERGTREcG89PjN9A7pho2u8Ebv+zi3knrOJmqGcavhcKNiIiIiQK8PfhgYHNev6MZ3h5Wlu89RY/3lzNny3Eq2D0/JUbhRkRExGQWi4WBbWswZ9QN1A/z51RaFo9P3cTgz9ay90Sq2eW5HIUbERERJ1E/zDHD+FNd6+PlbmX1gdP0eH85Y3/eSVpWrtnluQyFGxERESfi7eHG453rsfCpjnRpFEau3WD8sgN0eXspP+pSVZEo3IiIiDihqEq+fH5va764tzU1KvmSkJLJY1M3MeTztexL1KWqwijciIiIOLHOjcL49cmbeLKL41LVqv2n6f7ecsbO20m6LlVdlsKNiIiIk/P2cOOJLvVY8GRHujQKdVyqWnqAzm8v5aet8bpUdRGFGxERERdRo7Ivn9/bhs+Htiaqkg8JKZmMnLKRv32xjn2JaWaX5zQUbkRERFxMl8ZhLHiyI090roenu5UV+07R4/1lvD5vFxnZulSlcCMiIuKCvD3ceLJrfRY86ZiIM8dm8OnS/XR5eyk/b6vYl6oUbkRERFxYzcp+TBzmuFQVGeLD8eRMRnyzkaET17H/ZMW8VKVwIyIiUg50aRzGwqc68vj5S1XL956i+3vL+O8vFe9SlcKNiIhIOeHt4cZT5y9V3dKgKjk2g0+WOC5V/bK94lyqUrgREREpZy5cqvpsaGuqBzsuVT3y9UbunbSeAxXgUpXFqCgx7ryUlBSCgoJITk4mMDDQ7HJERERK1blsG+OW7OPTpQfIttnxdLNy3w21ePimOlTy8zS7vCIrzve3wo2IiEgFcOhUOi//uIMlu08C4OPhxj3tazD8xusIDfQ2uborU7gphMKNiIhUVIZh8NuuRN5ftJetR5MB8HS3MqhNFA91rEP1YB+TKyyYwk0hFG5ERKSiMwyDpXtO8uFv+9hw+CwAHm4W+reM5NGb61Czsp/JFV5K4aYQCjciIiIOhmGw5sAZPvxtL6v2nwbAzWqhb0w1RtxSh7qhASZX+CeFm0Io3IiIiFxqw+EzfPTbPhafH5NjsUDPphGM6lSXRhHmf18q3BRC4UZERKRg244m89HivczfcSJvXZdGYTzWqS4xUcGm1aVwUwiFGxERkSvbnZDKR4v3MXfrcS4khZvqV+WxTnVpU6tSmdejcFMIhRsREZGi238yjXFL9jNz0zFsdkdkaFe7Eo93rkeHOpWxWCxlUofCTSEUbkRERIrvyJkMxi3dz7Tfj5Bjc0SHFjWCeaxTXW5pEFrqIUfhphAKNyIiIlcvPvkc45ceYOq6OLJy7QA0qRbIY53qcmvjcKzW0gk5CjeFULgRERG5didTs/h8xQG+Wn2YjGwbAPVC/RnVqS63RVfDrYRDjsJNIRRuRERESs7Z9GwmrTzIpFWHSM3MBRwh56fHb8TTveTm5y7O97dmBRcREZGrFuLnyVO3NmDlPzrxzK31CfH1oGWNkBINNsWlnhsREREpMelZuWTm2Kjs71Wi+y3O97d7iX6yiIiIVGh+Xu74eZkbL3RZSkRERMoVhRsREREpVxRuREREpFxRuBEREZFyxdRws2zZMnr37k21atWwWCzMmjXritssWbKEli1b4uXlRd26dZk8eXKp1ykiIiKuw9Rwk56eTkxMDB9//HGR2h88eJBevXpxyy23sHnzZkaPHs2DDz7I/PnzS7lSERERcRWm3qvVo0cPevToUeT2n376KbVr1+btt98GoFGjRqxYsYJ3332Xbt26lVaZIiIi4kJcaszN6tWr6dKlS7513bp1Y/Xq1QVuk5WVRUpKSr5FREREyi+XCjcJCQmEhYXlWxcWFkZKSgrnzp277DZjx44lKCgob4mKiiqLUkVERMQkLhVursaYMWNITk7OW44cOWJ2SSIiIlKKXGr6hfDwcE6cOJFv3YkTJwgMDMTHx+ey23h5eeHlVbLzW4iIiIjzcqmem9jYWBYtWpRv3YIFC4iNjTWpIhEREXE2poabtLQ0Nm/ezObNmwHHrd6bN28mLi4OcFxSGjp0aF77Rx55hAMHDvD3v/+dXbt28cknn/D999/z5JNPmlG+iIiIOCFTL0v9/vvv3HLLLXmvn3rqKQDuvfdeJk+eTHx8fF7QAahduzY//fQTTz75JO+//z6RkZF8/vnnxboN3DAMAN01JSIi4kIufG9f+B4vjMUoSqty5OjRo7pjSkRExEUdOXKEyMjIQttUuHBjt9s5fvw4AQEBWCyWEt13SkoKUVFRHDlyhMDAwBLdt7PRsZZfFel4dazlV0U63opyrIZhkJqaSrVq1bBaCx9V41J3S5UEq9V6xcR3rQIDA8v1f2B/pWMtvyrS8epYy6+KdLwV4ViDgoKK1M6l7pYSERERuRKFGxERESlXFG5KkJeXFy+99FKFeGigjrX8qkjHq2MtvyrS8VakYy2qCjegWERERMo39dyIiIhIuaJwIyIiIuWKwo2IiIiUKwo3IiIiUq4o3BTTxx9/TK1atfD29qZdu3asW7eu0PbTpk2jYcOGeHt706xZM37++ecyqvTqjR07ljZt2hAQEEBoaCj9+vVj9+7dhW4zefJkLBZLvsXb27uMKr42L7/88iW1N2zYsNBtXPG8AtSqVeuSY7VYLIwcOfKy7V3pvC5btozevXtTrVo1LBYLs2bNyve+YRi8+OKLRERE4OPjQ5cuXdi7d+8V91vc3/myUtjx5uTk8Nxzz9GsWTP8/PyoVq0aQ4cO5fjx44Xu82p+F8rClc7tsGHDLqm7e/fuV9yvM57bKx3r5X5/LRYLb775ZoH7dNbzWpoUborhu+++46mnnuKll15i48aNxMTE0K1bNxITEy/bftWqVQwaNIgHHniATZs20a9fP/r168f27dvLuPLiWbp0KSNHjmTNmjUsWLCAnJwcbr31VtLT0wvdLjAwkPj4+Lzl8OHDZVTxtWvSpEm+2lesWFFgW1c9rwDr16/Pd5wLFiwA4M477yxwG1c5r+np6cTExPDxxx9f9v3//ve/fPDBB3z66aesXbsWPz8/unXrRmZmZoH7LO7vfFkq7HgzMjLYuHEjL7zwAhs3buSHH35g9+7d9OnT54r7Lc7vQlm50rkF6N69e766p06dWug+nfXcXulY/3qM8fHxTJw4EYvFQv/+/QvdrzOe11JlSJG1bdvWGDlyZN5rm81mVKtWzRg7duxl2991111Gr1698q1r166d8fDDD5dqnSUtMTHRAIylS5cW2GbSpElGUFBQ2RVVgl566SUjJiamyO3Ly3k1DMN44oknjDp16hh2u/2y77vqeQWMmTNn5r222+1GeHi48eabb+atS0pKMry8vIypU6cWuJ/i/s6b5eLjvZx169YZgHH48OEC2xT3d8EMlzvWe++91+jbt2+x9uMK57Yo57Vv375Gp06dCm3jCue1pKnnpoiys7PZsGEDXbp0yVtntVrp0qULq1evvuw2q1evztceoFu3bgW2d1bJyckAVKpUqdB2aWlp1KxZk6ioKPr27cuOHTvKorwSsXfvXqpVq8Z1113HkCFDiIuLK7BteTmv2dnZfP3119x///2FTiLryuf1goMHD5KQkJDvvAUFBdGuXbsCz9vV/M47s+TkZCwWC8HBwYW2K87vgjNZsmQJoaGhNGjQgEcffZTTp08X2La8nNsTJ07w008/8cADD1yxraue16ulcFNEp06dwmazERYWlm99WFgYCQkJl90mISGhWO2dkd1uZ/To0Vx//fU0bdq0wHYNGjRg4sSJzJ49m6+//hq73U6HDh04evRoGVZ7ddq1a8fkyZP55ZdfGDduHAcPHuTGG28kNTX1su3Lw3kFmDVrFklJSQwbNqzANq58Xv/qwrkpznm7mt95Z5WZmclzzz3HoEGDCp1Ysbi/C86ie/fu/O9//2PRokW88cYbLF26lB49emCz2S7bvryc2y+//JKAgADuuOOOQtu56nm9FhVuVnApnpEjR7J9+/YrXp+NjY0lNjY273WHDh1o1KgR48eP59VXXy3tMq9Jjx498v4eHR1Nu3btqFmzJt9//32R/kXkqr744gt69OhBtWrVCmzjyudVHHJycrjrrrswDINx48YV2tZVfxcGDhyY9/dmzZoRHR1NnTp1WLJkCZ07dzaxstI1ceJEhgwZcsVB/q56Xq+Fem6KqEqVKri5uXHixIl860+cOEF4ePhltwkPDy9We2czatQo5s6dy+LFi4mMjCzWth4eHrRo0YJ9+/aVUnWlJzg4mPr16xdYu6ufV4DDhw+zcOFCHnzwwWJt56rn9cK5Kc55u5rfeWdzIdgcPnyYBQsWFNprczlX+l1wVtdddx1VqlQpsO7ycG6XL1/O7t27i/07DK57XotD4aaIPD09adWqFYsWLcpbZ7fbWbRoUb5/2f5VbGxsvvYACxYsKLC9szAMg1GjRjFz5kx+++03ateuXex92Gw2tm3bRkRERClUWLrS0tLYv39/gbW76nn9q0mTJhEaGkqvXr2KtZ2rntfatWsTHh6e77ylpKSwdu3aAs/b1fzOO5MLwWbv3r0sXLiQypUrF3sfV/pdcFZHjx7l9OnTBdbt6ucWHD2vrVq1IiYmptjbuup5LRazRzS7km+//dbw8vIyJk+ebPzxxx/GQw89ZAQHBxsJCQmGYRjG3/72N+Mf//hHXvuVK1ca7u7uxltvvWXs3LnTeOmllwwPDw9j27ZtZh1CkTz66KNGUFCQsWTJEiM+Pj5vycjIyGtz8bG+8sorxvz58439+/cbGzZsMAYOHGh4e3sbO3bsMOMQiuXpp582lixZYhw8eNBYuXKl0aVLF6NKlSpGYmKiYRjl57xeYLPZjBo1ahjPPffcJe+58nlNTU01Nm3aZGzatMkAjHfeecfYtGlT3t1Br7/+uhEcHGzMnj3b2Lp1q9G3b1+jdu3axrlz5/L20alTJ+PDDz/Me32l33kzFXa82dnZRp8+fYzIyEhj8+bN+X6Ps7Ky8vZx8fFe6XfBLIUda2pqqvHMM88Yq1evNg4ePGgsXLjQaNmypVGvXj0jMzMzbx+ucm6v9N+xYRhGcnKy4evra4wbN+6y+3CV81qaFG6K6cMPPzRq1KhheHp6Gm3btjXWrFmT917Hjh2Ne++9N1/777//3qhfv77h6elpNGnSxPjpp5/KuOLiAy67TJo0Ka/Nxcc6evTovJ9LWFiY0bNnT2Pjxo1lX/xVuPvuu42IiAjD09PTqF69unH33Xcb+/bty3u/vJzXC+bPn28Axu7duy95z5XP6+LFiy/73+2F47Hb7cYLL7xghIWFGV5eXkbnzp0v+RnUrFnTeOmll/KtK+x33kyFHe/BgwcL/D1evHhx3j4uPt4r/S6YpbBjzcjIMG699VajatWqhoeHh1GzZk1j+PDhl4QUVzm3V/rv2DAMY/z48YaPj4+RlJR02X24ynktTRbDMIxS7RoSERERKUMacyMiIiLlisKNiIiIlCsKNyIiIlKuKNyIiIhIuaJwIyIiIuWKwo2IiIiUKwo3IiIiUq4o3IhIhWexWJg1a5bZZYhICVG4ERFTDRs2DIvFcsnSvXt3s0sTERflbnYBIiLdu3dn0qRJ+dZ5eXmZVI2IuDr13IiI6by8vAgPD8+3hISEAI5LRuPGjaNHjx74+Phw3XXXMX369Hzbb9u2jU6dOuHj40PlypV56KGHSEtLy9dm4sSJNGnSBC8vLyIiIhg1alS+90+dOsXtt9+Or68v9erVY86cOaV70CJSahRuRMTpvfDCC/Tv358tW7YwZMgQBg4cyM6dOwFIT0+nW7duhISEsH79eqZNm8bChQvzhZdx48YxcuRIHnroIbZt28acOXOoW7duvs945ZVXuOuuu9i6dSs9e/ZkyJAhnDlzpkyPU0RKiNkzd4pIxXbvvfcabm5uhp+fX77ltddeMwzDMUv9I488km+bdu3aGY8++qhhGIYxYcIEIyQkxEhLS8t7/6effjKsVmvezNDVqlUznn/++QJrAIz/+7//y3udlpZmAMa8efNK7DhFpOxozI2ImO6WW25h3Lhx+dZVqlQp7++xsbH53ouNjWXz5s0A7Ny5k5iYGPz8/PLev/7667Hb7ezevRuLxcLx48fp3LlzoTVER0fn/d3Pz4/AwEASExOv9pBExEQKNyJiOj8/v0suE5UUHx+fIrXz8PDI99pisWC320ujJBEpZRpzIyJOb82aNZe8btSoEQCNGjViy5YtpKen572/cuVKrFYrDRo0ICAggFq1arFo0aIyrVlEzKOeGxExXVZWFgkJCfnWubu7U6VKFQCmTZtG69atueGGG/jmm29Yt24dX3zxBQBDhgzhpZde4t577+Xll1/m5MmTPPbYY/ztb38jLCwMgJdffplHHnmE0NBQevToQWpqKitXruSxxx4r2wMVkTKhcCMipvvll1+IiIjIt65Bgwbs2rULcNzJ9O233zJixAgiIiKYOnUqjRs3BsDX15f58+fzxBNP0KZNG3x9fenfvz/vvPNO3r7uvfdeMjMzeffdd3nmmWeoUqUKAwYMKLsDFJEyZTEMwzC7CBGRglgsFmbOnEm/fv3MLkVEXITG3IiIiEi5onAjIiIi5YrG3IiIU9OVcxEpLvXciIiISLmicCMiIiLlisKNiIiIlCsKNyIiIlKuKNyIiIhIuaJwIyIiIuWKwo2IiIiUKwo3IiIiUq4o3IiIiEi58v+gncoo0hOQ9AAAAABJRU5ErkJggg==\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "model_3.save_weights('model_3_weights.h5')"
      ],
      "metadata": {
        "id": "xXcs4LS6v6C-"
      },
      "execution_count": 22,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "model_3.save('model_3.h5')"
      ],
      "metadata": {
        "id": "r5w2uOqlwDS8"
      },
      "execution_count": 23,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Load the model from the H5 file\n",
        "model_3.load_weights('/content/model_3_weights.h5')"
      ],
      "metadata": {
        "id": "mwN7q5mVWwXf"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Create a new model that outputs the fc7 features\n",
        "fc7_layer = model_3.get_layer('dense_7')\n",
        "fc7_output = fc7_layer.output\n",
        "model_3_withoudHead = Model(inputs=model_3.input, outputs=fc7_output)"
      ],
      "metadata": {
        "id": "RmgTW_ttqTIi"
      },
      "execution_count": 24,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Define the paths to the validation dataset\n",
        "test_dir = '/content/process_data/test'\n",
        "\n",
        "# Define the batch size for validation data\n",
        "batch_size = 1\n",
        "\n",
        "# Create an empty list to hold the feature vectors for the validation images\n",
        "test_features3 = []\n",
        "\n",
        "# Loop through the validation image directories\n",
        "for folder_name in os.listdir(test_dir):\n",
        "\n",
        "    test_directory = os.path.join(test_dir, folder_name)\n",
        "\n",
        "    # Create an empty list to hold the feature vectors for the validation images in the current directory\n",
        "    directory_features = []\n",
        "    \n",
        "    # Loop through the validation images in the current directory\n",
        "    for i in range(11, 16):\n",
        "        # Load the validation image using PIL\n",
        "        img_path = os.path.join(test_directory, f'{folder_name}_{i}.jpg')\n",
        "        img = image.load_img(img_path, target_size=(227, 227))\n",
        "        \n",
        "        # Convert the image to a numpy array and preprocess it\n",
        "        x = image.img_to_array(img)\n",
        "        x = np.expand_dims(x, axis=0)\n",
        "        x = preprocess_input(x)\n",
        "        \n",
        "        # Extract the features using the VGG19 model\n",
        "        features = model_3_withoudHead.predict(x)\n",
        "        \n",
        "        # Add the features to the list of features for the current directory\n",
        "        directory_features.append(features.flatten())\n",
        "    \n",
        "    # Convert the list of features for the current directory to a numpy array\n",
        "    directory_features = np.array(directory_features)\n",
        "    \n",
        "    # Add the numpy array to the list of feature vectors for all validation images\n",
        "    test_features3.append(directory_features)\n",
        "\n",
        "# Convert the list of feature vectors for all validation images to a numpy array\n",
        "test_features3 = np.array(test_features3)\n",
        "print(test_features3)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Sa50lEicqeh1",
        "outputId": "ca3944ac-0c3e-4f48-ff9e-6cca0559b4d7"
      },
      "execution_count": 26,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 757ms/step\n",
            "1/1 [==============================] - 1s 633ms/step\n",
            "1/1 [==============================] - 1s 656ms/step\n",
            "1/1 [==============================] - 1s 750ms/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 702ms/step\n",
            "1/1 [==============================] - 1s 659ms/step\n",
            "1/1 [==============================] - 1s 632ms/step\n",
            "1/1 [==============================] - 1s 667ms/step\n",
            "1/1 [==============================] - 1s 638ms/step\n",
            "1/1 [==============================] - 1s 652ms/step\n",
            "1/1 [==============================] - 1s 660ms/step\n",
            "1/1 [==============================] - 1s 664ms/step\n",
            "1/1 [==============================] - 1s 625ms/step\n",
            "1/1 [==============================] - 1s 658ms/step\n",
            "1/1 [==============================] - 1s 653ms/step\n",
            "1/1 [==============================] - 1s 640ms/step\n",
            "1/1 [==============================] - 1s 636ms/step\n",
            "1/1 [==============================] - 1s 645ms/step\n",
            "1/1 [==============================] - 1s 897ms/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 671ms/step\n",
            "1/1 [==============================] - 1s 649ms/step\n",
            "1/1 [==============================] - 1s 674ms/step\n",
            "1/1 [==============================] - 1s 641ms/step\n",
            "1/1 [==============================] - 1s 657ms/step\n",
            "1/1 [==============================] - 1s 625ms/step\n",
            "1/1 [==============================] - 1s 666ms/step\n",
            "1/1 [==============================] - 1s 644ms/step\n",
            "1/1 [==============================] - 1s 653ms/step\n",
            "1/1 [==============================] - 1s 641ms/step\n",
            "1/1 [==============================] - 1s 640ms/step\n",
            "1/1 [==============================] - 1s 643ms/step\n",
            "1/1 [==============================] - 1s 650ms/step\n",
            "1/1 [==============================] - 1s 672ms/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 908ms/step\n",
            "1/1 [==============================] - 1s 659ms/step\n",
            "1/1 [==============================] - 1s 647ms/step\n",
            "1/1 [==============================] - 1s 645ms/step\n",
            "1/1 [==============================] - 1s 631ms/step\n",
            "1/1 [==============================] - 1s 672ms/step\n",
            "1/1 [==============================] - 1s 647ms/step\n",
            "1/1 [==============================] - 1s 665ms/step\n",
            "1/1 [==============================] - 1s 658ms/step\n",
            "1/1 [==============================] - 1s 664ms/step\n",
            "1/1 [==============================] - 1s 642ms/step\n",
            "1/1 [==============================] - 1s 632ms/step\n",
            "1/1 [==============================] - 1s 622ms/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 825ms/step\n",
            "1/1 [==============================] - 1s 627ms/step\n",
            "1/1 [==============================] - 1s 629ms/step\n",
            "1/1 [==============================] - 1s 648ms/step\n",
            "1/1 [==============================] - 1s 625ms/step\n",
            "1/1 [==============================] - 1s 665ms/step\n",
            "1/1 [==============================] - 1s 644ms/step\n",
            "1/1 [==============================] - 1s 670ms/step\n",
            "1/1 [==============================] - 2s 2s/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 646ms/step\n",
            "1/1 [==============================] - 1s 655ms/step\n",
            "1/1 [==============================] - 1s 634ms/step\n",
            "1/1 [==============================] - 1s 655ms/step\n",
            "1/1 [==============================] - 1s 635ms/step\n",
            "1/1 [==============================] - 1s 660ms/step\n",
            "1/1 [==============================] - 1s 627ms/step\n",
            "1/1 [==============================] - 1s 660ms/step\n",
            "1/1 [==============================] - 1s 654ms/step\n",
            "1/1 [==============================] - 1s 652ms/step\n",
            "1/1 [==============================] - 1s 644ms/step\n",
            "1/1 [==============================] - 1s 645ms/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 933ms/step\n",
            "1/1 [==============================] - 1s 655ms/step\n",
            "1/1 [==============================] - 1s 663ms/step\n",
            "1/1 [==============================] - 1s 646ms/step\n",
            "1/1 [==============================] - 1s 642ms/step\n",
            "1/1 [==============================] - 1s 627ms/step\n",
            "1/1 [==============================] - 1s 623ms/step\n",
            "1/1 [==============================] - 1s 643ms/step\n",
            "1/1 [==============================] - 1s 638ms/step\n",
            "1/1 [==============================] - 1s 655ms/step\n",
            "1/1 [==============================] - 1s 641ms/step\n",
            "1/1 [==============================] - 1s 665ms/step\n",
            "1/1 [==============================] - 1s 632ms/step\n",
            "1/1 [==============================] - 1s 671ms/step\n",
            "1/1 [==============================] - 1s 715ms/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 770ms/step\n",
            "1/1 [==============================] - 1s 671ms/step\n",
            "1/1 [==============================] - 1s 644ms/step\n",
            "1/1 [==============================] - 1s 650ms/step\n",
            "1/1 [==============================] - 1s 650ms/step\n",
            "1/1 [==============================] - 1s 623ms/step\n",
            "1/1 [==============================] - 1s 657ms/step\n",
            "1/1 [==============================] - 1s 626ms/step\n",
            "1/1 [==============================] - 1s 624ms/step\n",
            "1/1 [==============================] - 1s 640ms/step\n",
            "1/1 [==============================] - 1s 640ms/step\n",
            "1/1 [==============================] - 1s 644ms/step\n",
            "1/1 [==============================] - 1s 652ms/step\n",
            "1/1 [==============================] - 1s 699ms/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 740ms/step\n",
            "1/1 [==============================] - 1s 648ms/step\n",
            "1/1 [==============================] - 1s 662ms/step\n",
            "1/1 [==============================] - 1s 627ms/step\n",
            "1/1 [==============================] - 1s 628ms/step\n",
            "1/1 [==============================] - 1s 650ms/step\n",
            "1/1 [==============================] - 1s 647ms/step\n",
            "1/1 [==============================] - 1s 628ms/step\n",
            "1/1 [==============================] - 1s 642ms/step\n",
            "1/1 [==============================] - 1s 629ms/step\n",
            "1/1 [==============================] - 1s 630ms/step\n",
            "1/1 [==============================] - 1s 643ms/step\n",
            "1/1 [==============================] - 1s 652ms/step\n",
            "1/1 [==============================] - 1s 637ms/step\n",
            "1/1 [==============================] - 1s 753ms/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 877ms/step\n",
            "1/1 [==============================] - 1s 630ms/step\n",
            "1/1 [==============================] - 1s 635ms/step\n",
            "1/1 [==============================] - 1s 638ms/step\n",
            "1/1 [==============================] - 1s 632ms/step\n",
            "1/1 [==============================] - 1s 630ms/step\n",
            "1/1 [==============================] - 1s 631ms/step\n",
            "1/1 [==============================] - 1s 645ms/step\n",
            "1/1 [==============================] - 1s 622ms/step\n",
            "1/1 [==============================] - 1s 634ms/step\n",
            "1/1 [==============================] - 1s 641ms/step\n",
            "1/1 [==============================] - 1s 623ms/step\n",
            "1/1 [==============================] - 1s 639ms/step\n",
            "1/1 [==============================] - 1s 650ms/step\n",
            "1/1 [==============================] - 1s 635ms/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 901ms/step\n",
            "1/1 [==============================] - 1s 632ms/step\n",
            "1/1 [==============================] - 1s 639ms/step\n",
            "1/1 [==============================] - 1s 645ms/step\n",
            "1/1 [==============================] - 1s 633ms/step\n",
            "1/1 [==============================] - 1s 648ms/step\n",
            "1/1 [==============================] - 1s 653ms/step\n",
            "1/1 [==============================] - 1s 630ms/step\n",
            "1/1 [==============================] - 1s 649ms/step\n",
            "1/1 [==============================] - 1s 657ms/step\n",
            "1/1 [==============================] - 1s 620ms/step\n",
            "1/1 [==============================] - 1s 643ms/step\n",
            "1/1 [==============================] - 1s 643ms/step\n",
            "1/1 [==============================] - 1s 637ms/step\n",
            "1/1 [==============================] - 1s 690ms/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 820ms/step\n",
            "1/1 [==============================] - 1s 655ms/step\n",
            "1/1 [==============================] - 1s 639ms/step\n",
            "1/1 [==============================] - 1s 652ms/step\n",
            "1/1 [==============================] - 1s 651ms/step\n",
            "1/1 [==============================] - 1s 698ms/step\n",
            "1/1 [==============================] - 1s 653ms/step\n",
            "1/1 [==============================] - 1s 690ms/step\n",
            "1/1 [==============================] - 1s 649ms/step\n",
            "1/1 [==============================] - 1s 638ms/step\n",
            "1/1 [==============================] - 1s 643ms/step\n",
            "1/1 [==============================] - 1s 661ms/step\n",
            "1/1 [==============================] - 1s 648ms/step\n",
            "1/1 [==============================] - 1s 626ms/step\n",
            "1/1 [==============================] - 1s 876ms/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 634ms/step\n",
            "1/1 [==============================] - 1s 646ms/step\n",
            "1/1 [==============================] - 1s 658ms/step\n",
            "1/1 [==============================] - 1s 664ms/step\n",
            "1/1 [==============================] - 1s 672ms/step\n",
            "1/1 [==============================] - 1s 634ms/step\n",
            "1/1 [==============================] - 1s 641ms/step\n",
            "1/1 [==============================] - 1s 700ms/step\n",
            "1/1 [==============================] - 1s 675ms/step\n",
            "1/1 [==============================] - 1s 671ms/step\n",
            "1/1 [==============================] - 1s 641ms/step\n",
            "1/1 [==============================] - 1s 644ms/step\n",
            "1/1 [==============================] - 1s 648ms/step\n",
            "1/1 [==============================] - 1s 622ms/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 925ms/step\n",
            "1/1 [==============================] - 1s 610ms/step\n",
            "1/1 [==============================] - 1s 626ms/step\n",
            "1/1 [==============================] - 1s 662ms/step\n",
            "1/1 [==============================] - 1s 629ms/step\n",
            "1/1 [==============================] - 1s 647ms/step\n",
            "1/1 [==============================] - 1s 658ms/step\n",
            "1/1 [==============================] - 1s 692ms/step\n",
            "1/1 [==============================] - 1s 657ms/step\n",
            "1/1 [==============================] - 1s 667ms/step\n",
            "1/1 [==============================] - 1s 662ms/step\n",
            "1/1 [==============================] - 1s 674ms/step\n",
            "1/1 [==============================] - 1s 675ms/step\n",
            "1/1 [==============================] - 1s 645ms/step\n",
            "1/1 [==============================] - 1s 759ms/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 829ms/step\n",
            "1/1 [==============================] - 1s 636ms/step\n",
            "1/1 [==============================] - 1s 667ms/step\n",
            "1/1 [==============================] - 1s 664ms/step\n",
            "1/1 [==============================] - 1s 638ms/step\n",
            "1/1 [==============================] - 1s 658ms/step\n",
            "1/1 [==============================] - 1s 647ms/step\n",
            "1/1 [==============================] - 1s 630ms/step\n",
            "1/1 [==============================] - 1s 643ms/step\n",
            "1/1 [==============================] - 1s 647ms/step\n",
            "1/1 [==============================] - 1s 633ms/step\n",
            "1/1 [==============================] - 1s 643ms/step\n",
            "1/1 [==============================] - 1s 634ms/step\n",
            "1/1 [==============================] - 1s 634ms/step\n",
            "1/1 [==============================] - 1s 752ms/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 687ms/step\n",
            "1/1 [==============================] - 1s 658ms/step\n",
            "1/1 [==============================] - 1s 639ms/step\n",
            "1/1 [==============================] - 1s 631ms/step\n",
            "1/1 [==============================] - 1s 626ms/step\n",
            "1/1 [==============================] - 1s 658ms/step\n",
            "1/1 [==============================] - 1s 634ms/step\n",
            "1/1 [==============================] - 1s 646ms/step\n",
            "1/1 [==============================] - 1s 640ms/step\n",
            "1/1 [==============================] - 1s 622ms/step\n",
            "1/1 [==============================] - 1s 628ms/step\n",
            "1/1 [==============================] - 1s 649ms/step\n",
            "1/1 [==============================] - 1s 622ms/step\n",
            "1/1 [==============================] - 1s 651ms/step\n",
            "1/1 [==============================] - 1s 842ms/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 1s/step\n",
            "1/1 [==============================] - 1s 635ms/step\n",
            "1/1 [==============================] - 1s 629ms/step\n",
            "[[[0.         0.         5.8425794  ... 0.         0.         5.759424  ]\n",
            "  [0.         0.         4.1405497  ... 0.         0.         3.165495  ]\n",
            "  [0.         0.         4.9139647  ... 0.         0.         2.4369574 ]\n",
            "  [0.         0.         2.6573274  ... 0.         0.         3.3722765 ]\n",
            "  [0.         0.         4.929103   ... 0.         0.         3.2213151 ]]\n",
            "\n",
            " [[0.         0.         3.4174056  ... 0.         0.         3.6706314 ]\n",
            "  [0.         0.         6.598283   ... 0.         0.         4.1776195 ]\n",
            "  [0.         0.         5.444682   ... 0.         2.320658   0.        ]\n",
            "  [0.         0.         6.518884   ... 0.         1.0204501  1.6027808 ]\n",
            "  [0.         0.         4.6642356  ... 0.         0.07766781 2.706942  ]]\n",
            "\n",
            " [[0.         0.         2.9137464  ... 0.         1.2121643  2.0550025 ]\n",
            "  [0.         0.         6.5099177  ... 0.         5.368968   0.        ]\n",
            "  [0.         0.         5.3355403  ... 0.         5.0574484  0.        ]\n",
            "  [0.         0.         4.1895785  ... 0.         3.4074247  0.        ]\n",
            "  [0.         0.         5.014804   ... 0.         5.8952613  0.        ]]\n",
            "\n",
            " ...\n",
            "\n",
            " [[0.         0.         2.5507073  ... 0.         1.1019531  1.3409367 ]\n",
            "  [0.         0.         3.1533163  ... 0.         1.8383882  1.7223758 ]\n",
            "  [0.         0.         2.8979046  ... 0.         1.6754642  0.5256036 ]\n",
            "  [0.         0.         2.8958895  ... 0.         0.93971163 0.9863464 ]\n",
            "  [0.         0.         3.229153   ... 0.         0.7024065  2.2291894 ]]\n",
            "\n",
            " [[0.         0.         8.885365   ... 0.         3.9265594  0.12261014]\n",
            "  [0.         0.         7.3906755  ... 0.         4.8613553  0.        ]\n",
            "  [0.         0.         6.5140443  ... 0.         5.2975307  0.        ]\n",
            "  [0.         0.         6.3561587  ... 0.         5.9723725  0.        ]\n",
            "  [0.         0.         7.102547   ... 0.         7.6105275  0.        ]]\n",
            "\n",
            " [[0.         0.         3.0768487  ... 0.         1.3538668  1.0312754 ]\n",
            "  [0.         0.         3.5620117  ... 0.         3.1782858  0.        ]\n",
            "  [0.         0.         3.9946032  ... 0.         0.88033265 1.2401446 ]\n",
            "  [0.         0.         4.3615813  ... 0.         0.9558677  1.5349827 ]\n",
            "  [0.         0.         3.618316   ... 0.         0.6498989  2.8429453 ]]]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Calculate the elementwise average\n",
        "average_features = np.mean([test_features, test_features2, test_features3], axis=0)\n",
        "\n",
        "print(average_features.shape)\n",
        "print(average_features)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "FR8w7p-AqfBM",
        "outputId": "ed66a703-ff54-4de4-f50e-ab63fc32c223"
      },
      "execution_count": 30,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(50, 5, 4096)\n",
            "[[[0.         4.713218   2.7642715  ... 0.         0.         1.919808  ]\n",
            "  [0.         4.0967665  2.114869   ... 0.         0.         1.0551649 ]\n",
            "  [0.         3.901858   2.85447    ... 0.         0.         0.8123191 ]\n",
            "  [0.         3.1763504  1.4915842  ... 0.         0.         1.1240922 ]\n",
            "  [0.         4.1090846  2.4074576  ... 0.         0.         1.0737717 ]]\n",
            "\n",
            " [[0.         4.312464   1.865963   ... 0.         0.         1.2235438 ]\n",
            "  [0.         5.7185884  3.7399063  ... 0.         0.         1.3925399 ]\n",
            "  [0.         5.6229606  3.68168    ... 0.         0.77355266 0.        ]\n",
            "  [0.         5.3756275  4.008633   ... 0.         0.34015003 0.5342603 ]\n",
            "  [0.         4.6479864  2.6824112  ... 0.         0.02588927 0.902314  ]]\n",
            "\n",
            " [[0.         3.6396341  1.74167    ... 0.         0.40405476 0.68500084]\n",
            "  [0.         4.880298   3.7578182  ... 0.         1.789656   0.        ]\n",
            "  [0.         3.9934275  2.9198897  ... 0.         1.6858162  0.        ]\n",
            "  [0.         3.9235609  2.4741728  ... 0.         1.1358082  0.        ]\n",
            "  [0.         4.181917   3.082695   ... 0.         1.965087   0.        ]]\n",
            "\n",
            " ...\n",
            "\n",
            " [[0.         2.9507687  1.4436048  ... 0.         0.3673177  0.4469789 ]\n",
            "  [0.         3.8976362  1.6862378  ... 0.         0.61279607 0.57412523]\n",
            "  [0.         3.3250558  1.8501936  ... 0.         0.5584881  0.17520119]\n",
            "  [0.         3.1017656  1.6061192  ... 0.         0.31323722 0.32878214]\n",
            "  [0.         3.435799   1.671986   ... 0.         0.23413551 0.74306315]]\n",
            "\n",
            " [[0.         6.938862   4.896547   ... 0.         1.3088531  0.04087005]\n",
            "  [0.         6.0237274  3.5746963  ... 0.         1.6204518  0.        ]\n",
            "  [0.         3.8961363  3.9435558  ... 0.         1.7658435  0.        ]\n",
            "  [0.         4.838141   4.1891007  ... 0.         1.9907908  0.        ]\n",
            "  [0.         4.480369   3.6028233  ... 0.         2.5368426  0.        ]]\n",
            "\n",
            " [[0.         3.712782   1.7882005  ... 0.         0.45128894 0.34375846]\n",
            "  [0.         4.1811576  2.0424907  ... 0.         1.0594286  0.        ]\n",
            "  [0.         3.99788    2.1030338  ... 0.         0.29344422 0.41338155]\n",
            "  [0.         4.3773026  2.200648   ... 0.         0.31862256 0.5116609 ]\n",
            "  [0.         4.799209   1.9225969  ... 0.         0.21663296 0.94764847]]]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Create a list for enrollment features for all subjects\n",
        "enrollment_features = [[0] * 4096 for i in range(50)]\n",
        "\n",
        "# Create a list for verification features for all subjects and each verification image\n",
        "verification_features = [[[0] * 4096 for k in range(4)] for j in range(50)]\n",
        "\n",
        "# Populate the enrollment_features list\n",
        "for i in range(0,50):\n",
        "    enrollment_features[i] = average_features[i][0]\n",
        "\n",
        "# Populate the verification_features list\n",
        "for j in range(0,50):\n",
        "    verification_features[j][0] = average_features[j][1]\n",
        "    verification_features[j][1] = average_features[j][2]\n",
        "    verification_features[j][2] = average_features[j][3]\n",
        "    verification_features[j][3] = average_features[j][4]\n",
        "\n",
        "# Create a numpy array to store the cosine similarities\n",
        "cosine_similarities_avg = np.zeros((10000, 2))\n",
        "\n",
        "# Iterate over each person and their enrollment and verification images\n",
        "for i in range(0,50):\n",
        "    a = 200 * i\n",
        "    for j in range(0,50):\n",
        "        b = a + (4 * j)\n",
        "        \n",
        "        # Get the enrollment feature and the four verification features\n",
        "        enrollment_features_2d = enrollment_features[i].reshape(1, -1)\n",
        "        verification_features_1_2d = verification_features[j][0].reshape(1, -1)\n",
        "        verification_features_2_2d = verification_features[j][1].reshape(1, -1)\n",
        "        verification_features_3_2d = verification_features[j][2].reshape(1, -1)\n",
        "        verification_features_4_2d = verification_features[j][3].reshape(1, -1)\n",
        "        # Calculate the cosine similarity between the enrollment and verification features\n",
        "        cosine_similarities_avg[b][0] = cosine_similarity(enrollment_features_2d, verification_features_1_2d)\n",
        "        cosine_similarities_avg[b+1][0] = cosine_similarity(enrollment_features_2d, verification_features_2_2d)\n",
        "        cosine_similarities_avg[b+2][0] = cosine_similarity(enrollment_features_2d, verification_features_3_2d)\n",
        "        cosine_similarities_avg[b+3][0] = cosine_similarity(enrollment_features_2d, verification_features_4_2d)\n",
        "        # Set the cosine similarity to 1 if the enrollment and verification images are from the same person\n",
        "        if i == j:\n",
        "            cosine_similarities_avg[b][1] = 1\n",
        "            cosine_similarities_avg[b+1][1] = 1\n",
        "            cosine_similarities_avg[b+2][1] = 1\n",
        "            cosine_similarities_avg[b+3][1] = 1\n",
        "\n",
        "print(cosine_similarities_avg)\n",
        "print(cosine_similarities_avg.shape)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "RQwlHSBzsPQH",
        "outputId": "77654ac5-ffd9-461a-b090-e5fcbd922b26"
      },
      "execution_count": 31,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[[0.99355602 1.        ]\n",
            " [0.98254967 1.        ]\n",
            " [0.99263692 1.        ]\n",
            " ...\n",
            " [0.9961496  1.        ]\n",
            " [0.99550915 1.        ]\n",
            " [0.99481201 1.        ]]\n",
            "(10000, 2)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Transpose and extract all cosine similarities\n",
        "cosine_similarities_t = np.transpose(cosine_similarities_avg)\n",
        "cosine_similarities_raw = cosine_similarities_t[0]\n",
        "\n",
        "# Plot the histogram\n",
        "plt.hist(cosine_similarities_raw, bins=50)\n",
        "plt.xlabel('Cosine similarity')\n",
        "plt.ylabel('Count')\n",
        "plt.title('Histogram of cosine similarities')\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 472
        },
        "id": "EyOPVnh5tJMj",
        "outputId": "03bf4079-3f29-461d-a7c8-d1510ff1f210"
      },
      "execution_count": 32,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Create emply list for genuine and imposter scores\n",
        "genuine_scores = []\n",
        "imposter_scores = []\n",
        "\n",
        "# Populate the lists based on true labels provided\n",
        "for i in range(cosine_similarities_avg.shape[0]):\n",
        "    if cosine_similarities_avg[i][1] == 1:\n",
        "        genuine_scores.append(cosine_similarities_avg[i][0])\n",
        "    else:\n",
        "        imposter_scores.append(cosine_similarities_avg[i][0])"
      ],
      "metadata": {
        "id": "JwaDlITQtreX"
      },
      "execution_count": 33,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# create arrays to store true labels and similarity scores\n",
        "true_labels = cosine_similarities_t[1]\n",
        "scores = cosine_similarities_t[0]\n",
        "\n",
        "# create arrays to store FPR and TPR for different thresholds\n",
        "fpr = []\n",
        "tpr = []\n",
        "thresholds = []\n",
        "\n",
        "# calculate the FPR and TPR for different thresholds\n",
        "fpr, tpr, thresholds = roc_curve(true_labels, scores)\n",
        "\n",
        "# calculate the area under the ROC curve (AUC)\n",
        "auc_score = auc(fpr, tpr)\n",
        "\n",
        "# calculate d prime\n",
        "d_prime = np.abs(np.mean(scores[true_labels == 1]) - np.mean(scores[true_labels == 0])) / np.sqrt(0.5 * (np.std(scores[true_labels == 1]) ** 2 + np.std(scores[true_labels == 0]) ** 2))\n",
        "\n",
        "# Calculate the optimal operating point on the ROC curve\n",
        "optimal_idx = np.argmax(tpr - fpr)\n",
        "OPTROCPT = (fpr[optimal_idx], tpr[optimal_idx])\n",
        "\n",
        "# plot the ROC curve\n",
        "plt.plot(fpr, tpr, color='darkorange', lw=2, label='ROC curve (AUC = %0.4f, d\\' = %0.4f)' % (auc_score, d_prime))\n",
        "plt.plot(OPTROCPT[0], OPTROCPT[1], 'ro', label='Optimal Operating Point (fpr = %0.4f, tpr = %0.4f)' % (fpr[optimal_idx], tpr[optimal_idx]))\n",
        "plt.plot([0, 1], [0, 1], color='navy', lw=2, linestyle='--')\n",
        "plt.xlim([0.0, 1.0])\n",
        "plt.ylim([0.0, 1.0])\n",
        "plt.xlabel('False Positive Rate')\n",
        "plt.ylabel('True Positive Rate')\n",
        "plt.title('Receiver operating characteristic')\n",
        "plt.legend(loc=\"lower right\")\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 472
        },
        "id": "4DKu3wsyuV0p",
        "outputId": "085b6f54-73b4-4005-fdbf-41700836f60f"
      },
      "execution_count": 34,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Plot histogram for genuine_scores\n",
        "plt.hist(genuine_scores, bins=50, label='Genuine Scores')\n",
        "plt.title('Histogram of Genuine Scores')\n",
        "plt.xlabel('Cosine Similarity')\n",
        "plt.ylabel('Frequency')\n",
        "plt.legend(loc='upper right')\n",
        "plt.show()\n",
        "\n",
        "# Plot histogram for imposter_scores\n",
        "plt.hist(imposter_scores, bins=50, label='Imposter Scores')\n",
        "plt.title('Histogram of Imposter Scores')\n",
        "plt.xlabel('Cosine Similarity')\n",
        "plt.ylabel('Frequency')\n",
        "plt.legend(loc='upper right')\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 927
        },
        "id": "fJtM2w41t_jF",
        "outputId": "2b9e6969-8552-4d8c-9a88-a37f94d75646"
      },
      "execution_count": 35,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Define paths\n",
        "original_dir = '/content/data/cropped_faces'\n",
        "train_dir_independent = '/content/process_data_independant/train'\n",
        "val_dir_independent = '/content/process_data_independant/validation'\n",
        "test_dir_independent = '/content/process_data_independant/test'\n",
        "\n",
        "# Create train and validation folders\n",
        "os.makedirs(train_dir_independent, exist_ok=True)\n",
        "os.makedirs(val_dir_independent, exist_ok=True)\n",
        "os.makedirs(test_dir_independent, exist_ok=True)\n",
        "\n",
        "# Define the number of images to put in each split\n",
        "train_count = 40\n",
        "test_count = 10\n",
        "\n",
        "# Loop through each person\n",
        "for i in range(1, 51):\n",
        "    person = \"s{:02d}\".format(i)\n",
        "    \n",
        "    if i <= train_count:\n",
        "        os.makedirs(os.path.join(train_dir_independent, person), exist_ok=True)\n",
        "        os.makedirs(os.path.join(val_dir_independent, person), exist_ok=True)\n",
        "        for j in range(1, 11):\n",
        "            image_name = \"{}_{:02d}.jpg\".format(person, j)\n",
        "            source_path = os.path.join(original_dir, image_name)\n",
        "            dest_path = os.path.join(train_dir_independent, person, image_name)\n",
        "            shutil.copy(source_path, dest_path)\n",
        "        for j in range(11, 16):\n",
        "            image_name = \"{}_{:02d}.jpg\".format(person, j)\n",
        "            source_path = os.path.join(original_dir, image_name)\n",
        "            dest_path = os.path.join(val_dir_independent, person, image_name)\n",
        "            shutil.copy(source_path, dest_path)\n",
        "    else:\n",
        "        os.makedirs(os.path.join(test_dir_independent, person), exist_ok=True)\n",
        "        for j in range(1, 16):\n",
        "            image_name = \"{}_{:02d}.jpg\".format(person, j)\n",
        "            source_path = os.path.join(original_dir, image_name)\n",
        "            dest_path = os.path.join(test_dir_independent, person, image_name)\n",
        "            shutil.copy(source_path, dest_path)"
      ],
      "metadata": {
        "id": "RIBV0Azjw0gQ"
      },
      "execution_count": 3,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Define data generators\n",
        "train_datagen = ImageDataGenerator(\n",
        "    rescale=1./255,\n",
        "    rotation_range=20,  # rotate images by up to 20 degrees\n",
        "    width_shift_range=0.2,  # shift images horizontally by up to 20% of the image width\n",
        "    height_shift_range=0.2,  # shift images vertically by up to 20% of the image height\n",
        "    shear_range=0.2,  # apply shear transformation with a shear angle of up to 20 degrees\n",
        "    zoom_range=0.2,  # apply zoom transformation by up to 20%\n",
        "    horizontal_flip=True,  # flip images horizontally\n",
        "    fill_mode='nearest'  # fill any gaps created by the above transformations with the nearest pixel\n",
        ")\n",
        "val_datagen = ImageDataGenerator(\n",
        "    rescale=1./255,\n",
        "    rotation_range=20,  # rotate images by up to 20 degrees\n",
        "    width_shift_range=0.2,  # shift images horizontally by up to 20% of the image width\n",
        "    height_shift_range=0.2,  # shift images vertically by up to 20% of the image height\n",
        "    shear_range=0.2,  # apply shear transformation with a shear angle of up to 20 degrees\n",
        "    zoom_range=0.2,  # apply zoom transformation by up to 20%\n",
        "    horizontal_flip=True,  # flip images horizontally\n",
        "    fill_mode='nearest'  # fill any gaps created by the above transformations with the nearest pixel\n",
        ")\n",
        "\n",
        "\n",
        "train_generator = train_datagen.flow_from_directory(\n",
        "    train_dir_independent,\n",
        "    target_size = (227,227),\n",
        "    batch_size=64,\n",
        "    class_mode='categorical')\n",
        "\n",
        "val_generator = val_datagen.flow_from_directory(\n",
        "    val_dir_independent,\n",
        "    target_size = (227,227),\n",
        "    batch_size=64,\n",
        "    class_mode='categorical')"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "bNAQef5pRwfE",
        "outputId": "c1ab668d-969c-4c17-80f7-65f393df4364"
      },
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Found 400 images belonging to 40 classes.\n",
            "Found 200 images belonging to 40 classes.\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Load VGG19 model without top layers\n",
        "base_model = VGG19(weights='imagenet', include_top=False)\n",
        "\n",
        "# Add top layers\n",
        "x = base_model.output\n",
        "x1 = GlobalAveragePooling2D()(x)\n",
        "x2 = Dense(4096, activation='relu')(x1) # Keep the fully connected layer with 4096 units\n",
        "x3 = Dense(4096, activation='relu')(x2) # Keep the fully connected layer with 4096 units\n",
        "predictions = Dense(40, activation='softmax')(x3) # Change to 40 units for 40-class classifier\n",
        "\n",
        "# Create a new model with the added top layers\n",
        "model_independent = Model(inputs=base_model.input, outputs=predictions)\n",
        "\n",
        "# Freeze base model layers\n",
        "for layer in base_model.layers:\n",
        "    layer.trainable = False\n",
        "\n",
        "# Compile the model\n",
        "model_independent.compile(optimizer='adam', loss='categorical_crossentropy', metrics=['accuracy'])\n",
        "\n",
        "# Print the model summary\n",
        "model_independent.summary()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ojhZdBWxR8h2",
        "outputId": "af53bcae-51a1-46e1-c7ae-bedfe0b060e8"
      },
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"model_3\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " input_4 (InputLayer)        [(None, None, None, 3)]   0         \n",
            "                                                                 \n",
            " block1_conv1 (Conv2D)       (None, None, None, 64)    1792      \n",
            "                                                                 \n",
            " block1_conv2 (Conv2D)       (None, None, None, 64)    36928     \n",
            "                                                                 \n",
            " block1_pool (MaxPooling2D)  (None, None, None, 64)    0         \n",
            "                                                                 \n",
            " block2_conv1 (Conv2D)       (None, None, None, 128)   73856     \n",
            "                                                                 \n",
            " block2_conv2 (Conv2D)       (None, None, None, 128)   147584    \n",
            "                                                                 \n",
            " block2_pool (MaxPooling2D)  (None, None, None, 128)   0         \n",
            "                                                                 \n",
            " block3_conv1 (Conv2D)       (None, None, None, 256)   295168    \n",
            "                                                                 \n",
            " block3_conv2 (Conv2D)       (None, None, None, 256)   590080    \n",
            "                                                                 \n",
            " block3_conv3 (Conv2D)       (None, None, None, 256)   590080    \n",
            "                                                                 \n",
            " block3_conv4 (Conv2D)       (None, None, None, 256)   590080    \n",
            "                                                                 \n",
            " block3_pool (MaxPooling2D)  (None, None, None, 256)   0         \n",
            "                                                                 \n",
            " block4_conv1 (Conv2D)       (None, None, None, 512)   1180160   \n",
            "                                                                 \n",
            " block4_conv2 (Conv2D)       (None, None, None, 512)   2359808   \n",
            "                                                                 \n",
            " block4_conv3 (Conv2D)       (None, None, None, 512)   2359808   \n",
            "                                                                 \n",
            " block4_conv4 (Conv2D)       (None, None, None, 512)   2359808   \n",
            "                                                                 \n",
            " block4_pool (MaxPooling2D)  (None, None, None, 512)   0         \n",
            "                                                                 \n",
            " block5_conv1 (Conv2D)       (None, None, None, 512)   2359808   \n",
            "                                                                 \n",
            " block5_conv2 (Conv2D)       (None, None, None, 512)   2359808   \n",
            "                                                                 \n",
            " block5_conv3 (Conv2D)       (None, None, None, 512)   2359808   \n",
            "                                                                 \n",
            " block5_conv4 (Conv2D)       (None, None, None, 512)   2359808   \n",
            "                                                                 \n",
            " block5_pool (MaxPooling2D)  (None, None, None, 512)   0         \n",
            "                                                                 \n",
            " global_average_pooling2d_3   (None, 512)              0         \n",
            " (GlobalAveragePooling2D)                                        \n",
            "                                                                 \n",
            " dense_9 (Dense)             (None, 4096)              2101248   \n",
            "                                                                 \n",
            " dense_10 (Dense)            (None, 4096)              16781312  \n",
            "                                                                 \n",
            " dense_11 (Dense)            (None, 40)                163880    \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 39,070,824\n",
            "Trainable params: 19,046,440\n",
            "Non-trainable params: 20,024,384\n",
            "_________________________________________________________________\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Train the model\n",
        "history_independent_1 = model_independent.fit(\n",
        "    train_generator,\n",
        "    epochs=20,  # specify the number of epochs for training\n",
        "    validation_data=val_generator\n",
        ")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "XCYOh5DGSfN0",
        "outputId": "514d89b1-5579-4e3b-a93c-899aacbee78b"
      },
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/20\n",
            "7/7 [==============================] - 351s 53s/step - loss: 4.2632 - accuracy: 0.0150 - val_loss: 3.6991 - val_accuracy: 0.0350\n",
            "Epoch 2/20\n",
            "7/7 [==============================] - 347s 52s/step - loss: 3.7105 - accuracy: 0.0175 - val_loss: 3.6729 - val_accuracy: 0.0400\n",
            "Epoch 3/20\n",
            "7/7 [==============================] - 309s 46s/step - loss: 3.6773 - accuracy: 0.0350 - val_loss: 3.6375 - val_accuracy: 0.0500\n",
            "Epoch 4/20\n",
            "7/7 [==============================] - 305s 45s/step - loss: 3.6142 - accuracy: 0.0600 - val_loss: 3.5443 - val_accuracy: 0.0950\n",
            "Epoch 5/20\n",
            "7/7 [==============================] - 351s 52s/step - loss: 3.5043 - accuracy: 0.0725 - val_loss: 3.3734 - val_accuracy: 0.0800\n",
            "Epoch 6/20\n",
            "7/7 [==============================] - 345s 52s/step - loss: 3.3692 - accuracy: 0.0925 - val_loss: 3.2912 - val_accuracy: 0.0800\n",
            "Epoch 7/20\n",
            "7/7 [==============================] - 344s 52s/step - loss: 3.2618 - accuracy: 0.0775 - val_loss: 3.2191 - val_accuracy: 0.0800\n",
            "Epoch 8/20\n",
            "7/7 [==============================] - 309s 46s/step - loss: 3.2147 - accuracy: 0.0775 - val_loss: 3.1312 - val_accuracy: 0.1050\n",
            "Epoch 9/20\n",
            "7/7 [==============================] - 348s 52s/step - loss: 3.1737 - accuracy: 0.1350 - val_loss: 3.1796 - val_accuracy: 0.0850\n",
            "Epoch 10/20\n",
            "7/7 [==============================] - 355s 54s/step - loss: 3.1886 - accuracy: 0.1075 - val_loss: 3.0780 - val_accuracy: 0.1450\n",
            "Epoch 11/20\n",
            "7/7 [==============================] - 347s 56s/step - loss: 3.1448 - accuracy: 0.1075 - val_loss: 3.2486 - val_accuracy: 0.1200\n",
            "Epoch 12/20\n",
            "7/7 [==============================] - 348s 52s/step - loss: 3.0296 - accuracy: 0.1175 - val_loss: 3.0132 - val_accuracy: 0.1250\n",
            "Epoch 13/20\n",
            "7/7 [==============================] - 347s 52s/step - loss: 2.9026 - accuracy: 0.1475 - val_loss: 2.8290 - val_accuracy: 0.2050\n",
            "Epoch 14/20\n",
            "7/7 [==============================] - 345s 52s/step - loss: 2.8225 - accuracy: 0.1650 - val_loss: 3.0373 - val_accuracy: 0.1450\n",
            "Epoch 15/20\n",
            "7/7 [==============================] - 349s 53s/step - loss: 2.6922 - accuracy: 0.2125 - val_loss: 2.8937 - val_accuracy: 0.1600\n",
            "Epoch 16/20\n",
            "7/7 [==============================] - 312s 46s/step - loss: 2.6449 - accuracy: 0.2200 - val_loss: 2.6584 - val_accuracy: 0.2350\n",
            "Epoch 17/20\n",
            "7/7 [==============================] - 321s 48s/step - loss: 2.5040 - accuracy: 0.2500 - val_loss: 2.6405 - val_accuracy: 0.2400\n",
            "Epoch 18/20\n",
            "7/7 [==============================] - 312s 46s/step - loss: 2.4117 - accuracy: 0.2700 - val_loss: 2.6173 - val_accuracy: 0.2300\n",
            "Epoch 19/20\n",
            "7/7 [==============================] - 309s 46s/step - loss: 2.3865 - accuracy: 0.2700 - val_loss: 2.8300 - val_accuracy: 0.1900\n",
            "Epoch 20/20\n",
            "7/7 [==============================] - 353s 57s/step - loss: 2.3256 - accuracy: 0.2900 - val_loss: 2.4926 - val_accuracy: 0.2400\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Plot the training and validation accuracy\n",
        "plt.plot(history_independent_1.history['accuracy'])\n",
        "plt.plot(history_independent_1.history['val_accuracy'])\n",
        "plt.title('Model Accuracy')\n",
        "plt.ylabel('Accuracy')\n",
        "plt.xlabel('Epoch')\n",
        "plt.legend(['train', 'val'], loc='upper left')\n",
        "plt.show()\n",
        "\n",
        "# Plot the training and validation loss\n",
        "plt.plot(history_independent_1.history['loss'])\n",
        "plt.plot(history_independent_1.history['val_loss'])\n",
        "plt.title('Model Loss')\n",
        "plt.ylabel('Loss')\n",
        "plt.xlabel('Epoch')\n",
        "plt.legend(['train', 'val'], loc='upper left')\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 927
        },
        "id": "FesxDnB24HW0",
        "outputId": "42150d1d-0c45-4f1f-b262-ae4cde6eeec2"
      },
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "model_independent.save_weights('model_independent_weights.h5')"
      ],
      "metadata": {
        "id": "ZIINui89XlBt"
      },
      "execution_count": 11,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "model_independent.save('model_independent.h5')"
      ],
      "metadata": {
        "id": "a3W8ILsoX7FI"
      },
      "execution_count": 12,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Create a new model that outputs the fc7 features\n",
        "fc7_layer = model_independent.get_layer('dense_10')\n",
        "fc7_output = fc7_layer.output\n",
        "model_independent_withoudHead = Model(inputs=model_independent.input, outputs=fc7_output)"
      ],
      "metadata": {
        "id": "1-p6SiDRYAcE"
      },
      "execution_count": 13,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Define the paths to the validation dataset\n",
        "test_dir = '/content/process_data_independant/test'\n",
        "\n",
        "# Define the batch size for validation data\n",
        "batch_size = 1\n",
        "\n",
        "# Create an empty list to hold the feature vectors for the validation images\n",
        "test_features_independent = []\n",
        "\n",
        "# Loop through the validation image directories\n",
        "for folder_name in os.listdir(test_dir):\n",
        "\n",
        "    test_directory = os.path.join(test_dir, folder_name)\n",
        "\n",
        "    # Create an empty list to hold the feature vectors for the validation images in the current directory\n",
        "    directory_features = []\n",
        "    \n",
        "    # Loop through the validation images in the current directory\n",
        "    for i in range(1, 16):\n",
        "        # Load the validation image using PIL\n",
        "        if i < 10:\n",
        "            img_path = os.path.join(test_directory, f'{folder_name}_0{i}.jpg')\n",
        "        else:\n",
        "            img_path = os.path.join(test_directory, f'{folder_name}_{i}.jpg')\n",
        "        img = image.load_img(img_path, target_size=(227, 227))\n",
        "        \n",
        "        # Convert the image to a numpy array and preprocess it\n",
        "        x = image.img_to_array(img)\n",
        "        x = np.expand_dims(x, axis=0)\n",
        "        x = preprocess_input(x)\n",
        "        \n",
        "        # Extract the features using the VGG19 model\n",
        "        features = model_independent_withoudHead.predict(x)\n",
        "        \n",
        "        # Add the features to the list of features for the current directory\n",
        "        directory_features.append(features.flatten())\n",
        "    \n",
        "    # Convert the list of features for the current directory to a numpy array\n",
        "    directory_features = np.array(directory_features)\n",
        "    \n",
        "    # Add the numpy array to the list of feature vectors for all validation images\n",
        "    test_features_independent.append(directory_features)\n",
        "\n",
        "# Convert the list of feature vectors for all validation images to a numpy array\n",
        "test_features_independent = np.array(test_features_independent)\n",
        "print(test_features_independent)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "q8Y04Kfra8nm",
        "outputId": "5ae53d90-be9f-4cd8-da4f-00ddb9062032"
      },
      "execution_count": 18,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1/1 [==============================] - 1s 516ms/step\n",
            "1/1 [==============================] - 1s 644ms/step\n",
            "1/1 [==============================] - 1s 719ms/step\n",
            "1/1 [==============================] - 1s 692ms/step\n",
            "1/1 [==============================] - 1s 568ms/step\n",
            "1/1 [==============================] - 0s 490ms/step\n",
            "1/1 [==============================] - 1s 501ms/step\n",
            "1/1 [==============================] - 1s 516ms/step\n",
            "1/1 [==============================] - 0s 498ms/step\n",
            "1/1 [==============================] - 0s 491ms/step\n",
            "1/1 [==============================] - 0s 480ms/step\n",
            "1/1 [==============================] - 0s 493ms/step\n",
            "1/1 [==============================] - 0s 484ms/step\n",
            "1/1 [==============================] - 0s 494ms/step\n",
            "1/1 [==============================] - 1s 527ms/step\n",
            "1/1 [==============================] - 1s 533ms/step\n",
            "1/1 [==============================] - 1s 549ms/step\n",
            "1/1 [==============================] - 0s 497ms/step\n",
            "1/1 [==============================] - 1s 532ms/step\n",
            "1/1 [==============================] - 0s 488ms/step\n",
            "1/1 [==============================] - 0s 491ms/step\n",
            "1/1 [==============================] - 1s 508ms/step\n",
            "1/1 [==============================] - 1s 555ms/step\n",
            "1/1 [==============================] - 1s 702ms/step\n",
            "1/1 [==============================] - 1s 692ms/step\n",
            "1/1 [==============================] - 1s 628ms/step\n",
            "1/1 [==============================] - 0s 484ms/step\n",
            "1/1 [==============================] - 1s 505ms/step\n",
            "1/1 [==============================] - 0s 495ms/step\n",
            "1/1 [==============================] - 0s 483ms/step\n",
            "1/1 [==============================] - 1s 503ms/step\n",
            "1/1 [==============================] - 0s 485ms/step\n",
            "1/1 [==============================] - 1s 512ms/step\n",
            "1/1 [==============================] - 0s 483ms/step\n",
            "1/1 [==============================] - 0s 481ms/step\n",
            "1/1 [==============================] - 0s 481ms/step\n",
            "1/1 [==============================] - 0s 493ms/step\n",
            "1/1 [==============================] - 1s 706ms/step\n",
            "1/1 [==============================] - 0s 485ms/step\n",
            "1/1 [==============================] - 1s 707ms/step\n",
            "1/1 [==============================] - 1s 509ms/step\n",
            "1/1 [==============================] - 0s 477ms/step\n",
            "1/1 [==============================] - 0s 485ms/step\n",
            "1/1 [==============================] - 1s 552ms/step\n",
            "1/1 [==============================] - 1s 700ms/step\n",
            "1/1 [==============================] - 1s 687ms/step\n",
            "1/1 [==============================] - 1s 590ms/step\n",
            "1/1 [==============================] - 0s 480ms/step\n",
            "1/1 [==============================] - 0s 495ms/step\n",
            "1/1 [==============================] - 0s 491ms/step\n",
            "1/1 [==============================] - 0s 487ms/step\n",
            "1/1 [==============================] - 0s 491ms/step\n",
            "1/1 [==============================] - 0s 480ms/step\n",
            "1/1 [==============================] - 0s 499ms/step\n",
            "1/1 [==============================] - 0s 487ms/step\n",
            "1/1 [==============================] - 0s 484ms/step\n",
            "1/1 [==============================] - 0s 487ms/step\n",
            "1/1 [==============================] - 0s 495ms/step\n",
            "1/1 [==============================] - 0s 486ms/step\n",
            "1/1 [==============================] - 1s 505ms/step\n",
            "1/1 [==============================] - 0s 489ms/step\n",
            "1/1 [==============================] - 0s 484ms/step\n",
            "1/1 [==============================] - 1s 506ms/step\n",
            "1/1 [==============================] - 0s 498ms/step\n",
            "1/1 [==============================] - 0s 491ms/step\n",
            "1/1 [==============================] - 1s 651ms/step\n",
            "1/1 [==============================] - 1s 707ms/step\n",
            "1/1 [==============================] - 1s 702ms/step\n",
            "1/1 [==============================] - 1s 534ms/step\n",
            "1/1 [==============================] - 1s 528ms/step\n",
            "1/1 [==============================] - 0s 490ms/step\n",
            "1/1 [==============================] - 0s 499ms/step\n",
            "1/1 [==============================] - 0s 484ms/step\n",
            "1/1 [==============================] - 0s 486ms/step\n",
            "1/1 [==============================] - 0s 487ms/step\n",
            "1/1 [==============================] - 1s 513ms/step\n",
            "1/1 [==============================] - 0s 500ms/step\n",
            "1/1 [==============================] - 1s 501ms/step\n",
            "1/1 [==============================] - 0s 490ms/step\n",
            "1/1 [==============================] - 0s 499ms/step\n",
            "1/1 [==============================] - 0s 490ms/step\n",
            "1/1 [==============================] - 0s 493ms/step\n",
            "1/1 [==============================] - 0s 492ms/step\n",
            "1/1 [==============================] - 1s 519ms/step\n",
            "1/1 [==============================] - 0s 481ms/step\n",
            "1/1 [==============================] - 0s 489ms/step\n",
            "1/1 [==============================] - 1s 509ms/step\n",
            "1/1 [==============================] - 1s 678ms/step\n",
            "1/1 [==============================] - 1s 696ms/step\n",
            "1/1 [==============================] - 1s 618ms/step\n",
            "1/1 [==============================] - 0s 479ms/step\n",
            "1/1 [==============================] - 1s 506ms/step\n",
            "1/1 [==============================] - 0s 484ms/step\n",
            "1/1 [==============================] - 0s 490ms/step\n",
            "1/1 [==============================] - 0s 468ms/step\n",
            "1/1 [==============================] - 0s 480ms/step\n",
            "1/1 [==============================] - 1s 501ms/step\n",
            "1/1 [==============================] - 0s 487ms/step\n",
            "1/1 [==============================] - 0s 493ms/step\n",
            "1/1 [==============================] - 0s 495ms/step\n",
            "1/1 [==============================] - 0s 485ms/step\n",
            "1/1 [==============================] - 1s 512ms/step\n",
            "1/1 [==============================] - 0s 488ms/step\n",
            "1/1 [==============================] - 1s 511ms/step\n",
            "1/1 [==============================] - 0s 498ms/step\n",
            "1/1 [==============================] - 0s 491ms/step\n",
            "1/1 [==============================] - 0s 487ms/step\n",
            "1/1 [==============================] - 0s 493ms/step\n",
            "1/1 [==============================] - 1s 589ms/step\n",
            "1/1 [==============================] - 1s 719ms/step\n",
            "1/1 [==============================] - 1s 695ms/step\n",
            "1/1 [==============================] - 1s 548ms/step\n",
            "1/1 [==============================] - 0s 477ms/step\n",
            "1/1 [==============================] - 0s 498ms/step\n",
            "1/1 [==============================] - 0s 498ms/step\n",
            "1/1 [==============================] - 0s 498ms/step\n",
            "1/1 [==============================] - 0s 493ms/step\n",
            "1/1 [==============================] - 0s 495ms/step\n",
            "1/1 [==============================] - 0s 482ms/step\n",
            "1/1 [==============================] - 0s 497ms/step\n",
            "1/1 [==============================] - 0s 496ms/step\n",
            "1/1 [==============================] - 0s 499ms/step\n",
            "1/1 [==============================] - 0s 491ms/step\n",
            "1/1 [==============================] - 1s 511ms/step\n",
            "1/1 [==============================] - 0s 488ms/step\n",
            "1/1 [==============================] - 0s 485ms/step\n",
            "1/1 [==============================] - 0s 481ms/step\n",
            "1/1 [==============================] - 0s 491ms/step\n",
            "1/1 [==============================] - 0s 491ms/step\n",
            "1/1 [==============================] - 0s 497ms/step\n",
            "1/1 [==============================] - 1s 673ms/step\n",
            "1/1 [==============================] - 1s 690ms/step\n",
            "1/1 [==============================] - 1s 651ms/step\n",
            "1/1 [==============================] - 0s 492ms/step\n",
            "1/1 [==============================] - 0s 485ms/step\n",
            "1/1 [==============================] - 1s 505ms/step\n",
            "1/1 [==============================] - 1s 513ms/step\n",
            "1/1 [==============================] - 0s 499ms/step\n",
            "1/1 [==============================] - 0s 499ms/step\n",
            "1/1 [==============================] - 1s 511ms/step\n",
            "1/1 [==============================] - 0s 478ms/step\n",
            "1/1 [==============================] - 0s 493ms/step\n",
            "1/1 [==============================] - 0s 492ms/step\n",
            "1/1 [==============================] - 1s 504ms/step\n",
            "1/1 [==============================] - 0s 486ms/step\n",
            "1/1 [==============================] - 0s 497ms/step\n",
            "1/1 [==============================] - 1s 513ms/step\n",
            "1/1 [==============================] - 0s 490ms/step\n",
            "1/1 [==============================] - 1s 511ms/step\n",
            "1/1 [==============================] - 1s 522ms/step\n",
            "[[[ 0.          0.          0.         ... 10.54312     0.08582778\n",
            "    0.        ]\n",
            "  [ 0.          0.          0.         ...  9.430253    0.5414949\n",
            "    0.        ]\n",
            "  [ 0.          0.          0.         ...  9.650424    0.\n",
            "    0.        ]\n",
            "  ...\n",
            "  [ 0.          0.          0.         ...  7.6797814   0.\n",
            "    0.        ]\n",
            "  [ 0.          0.          0.         ...  6.4337816   0.12662497\n",
            "    0.        ]\n",
            "  [ 0.          0.          0.         ...  7.3891854   0.8449629\n",
            "    0.        ]]\n",
            "\n",
            " [[ 0.          0.          0.         ...  7.0878296   0.\n",
            "    0.        ]\n",
            "  [ 0.          0.          0.         ...  6.0864573   0.08150569\n",
            "    0.        ]\n",
            "  [ 0.          0.          0.         ...  4.731443    0.\n",
            "    0.        ]\n",
            "  ...\n",
            "  [ 0.          0.          0.         ...  7.2884197   0.\n",
            "    0.        ]\n",
            "  [ 0.          0.          0.         ...  8.092626    0.42694688\n",
            "    0.        ]\n",
            "  [ 0.          0.          0.         ...  8.013678    0.\n",
            "    0.        ]]\n",
            "\n",
            " [[ 0.          0.          0.         ...  6.6392665   0.\n",
            "    0.        ]\n",
            "  [ 0.          0.          0.         ...  7.1452227   0.45462313\n",
            "    0.        ]\n",
            "  [ 0.          0.          0.         ...  5.7503786   0.01913347\n",
            "    0.        ]\n",
            "  ...\n",
            "  [ 0.          0.          0.         ...  6.506767    0.9863842\n",
            "    0.        ]\n",
            "  [ 0.          0.          0.         ...  5.311013    0.608117\n",
            "    0.        ]\n",
            "  [ 0.          0.          0.         ...  5.4132576   1.9614972\n",
            "    0.        ]]\n",
            "\n",
            " ...\n",
            "\n",
            " [[ 0.          0.          0.         ...  5.0430913   0.75247025\n",
            "    0.        ]\n",
            "  [ 0.          0.          0.         ...  5.864534    2.3749604\n",
            "    0.        ]\n",
            "  [ 0.          0.          0.         ...  5.310254    0.49169275\n",
            "    0.        ]\n",
            "  ...\n",
            "  [ 0.          0.          0.         ...  5.5599494   0.8267071\n",
            "    0.        ]\n",
            "  [ 0.          0.          0.         ...  5.1300597   1.6311343\n",
            "    0.        ]\n",
            "  [ 0.          0.          0.         ...  5.1319847   0.8539063\n",
            "    0.        ]]\n",
            "\n",
            " [[ 0.          0.          0.         ... 10.7840185   1.8049805\n",
            "    0.        ]\n",
            "  [ 0.          0.          0.         ...  4.8522363   0.53565896\n",
            "    0.        ]\n",
            "  [ 0.          0.          0.         ...  6.745107    0.23757239\n",
            "    0.        ]\n",
            "  ...\n",
            "  [ 0.          0.          0.         ...  5.2343254   0.850933\n",
            "    0.        ]\n",
            "  [ 0.          0.          0.         ...  5.939579    0.7474024\n",
            "    0.        ]\n",
            "  [ 0.          0.          0.         ...  5.247729    0.63993263\n",
            "    0.        ]]\n",
            "\n",
            " [[ 0.          0.          0.         ...  5.5242825   1.537335\n",
            "    0.        ]\n",
            "  [ 0.          0.          0.         ...  5.1568127   0.7982864\n",
            "    0.        ]\n",
            "  [ 0.          0.          0.         ...  3.510101    2.4322407\n",
            "    0.        ]\n",
            "  ...\n",
            "  [ 0.          0.          0.         ...  4.212257    1.9081255\n",
            "    0.        ]\n",
            "  [ 0.          0.          0.         ...  4.471874    1.9413598\n",
            "    0.        ]\n",
            "  [ 0.          0.          0.         ...  4.7638793   1.3632677\n",
            "    0.        ]]]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "print(test_features_independent.shape)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "mOmf49gcca_m",
        "outputId": "eb99b578-3360-47e1-d946-584bc7440160"
      },
      "execution_count": 19,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(10, 15, 4096)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Create a list for enrollment features for all subjects\n",
        "enrollment_features = [[0] * 4096 for i in range(10)]\n",
        "\n",
        "# Create a list for verification features for all subjects and each verification image\n",
        "verification_features = [[[0] * 4096 for k in range(14)] for j in range(10)]\n",
        "\n",
        "# Populate the enrollment_features list\n",
        "for i in range(0,10):\n",
        "    enrollment_features[i] = test_features_independent[i][0]\n",
        "\n",
        "# Populate the verification_features list\n",
        "for j in range(0,10):\n",
        "    for i in range(0,14):\n",
        "        verification_features[j][i] = test_features_independent[j][i+1]\n",
        "\n",
        "# Create a numpy array to store the cosine similarities\n",
        "cosine_similarities_independent = np.zeros((1400, 2))\n",
        "\n",
        "# Iterate over each person and their enrollment and verification images\n",
        "for i in range(0,10):\n",
        "    a = 140 * i\n",
        "    for j in range(0,10):\n",
        "        b = a + (14 * j)\n",
        "        enrollment_features_2d = enrollment_features[i].reshape(1, -1)\n",
        "        for k in range(0,14):\n",
        "            verification_features_2d = verification_features[j][k].reshape(1, -1)\n",
        "            cosine_similarities_independent[b+k][0] = cosine_similarity(enrollment_features_2d, verification_features_2d)\n",
        "        if i == j:\n",
        "            for k in range(0,14):\n",
        "                cosine_similarities_independent[b+k][1] = 1\n",
        "\n",
        "\n",
        "print(cosine_similarities_independent)\n",
        "print(cosine_similarities_independent.shape)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "kziQJfkPbtH0",
        "outputId": "ce7a1983-7614-439d-a497-cc8739727de3"
      },
      "execution_count": 25,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[[0.99686724 1.        ]\n",
            " [0.99397624 1.        ]\n",
            " [0.97055191 1.        ]\n",
            " ...\n",
            " [0.9673363  1.        ]\n",
            " [0.96765578 1.        ]\n",
            " [0.99450433 1.        ]]\n",
            "(1400, 2)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Transpose and extract all cosine similarities\n",
        "cosine_similarities_t = np.transpose(cosine_similarities_independent)\n",
        "cosine_similarities_raw = cosine_similarities_t[0]\n",
        "\n",
        "# Plot the histogram\n",
        "plt.hist(cosine_similarities_raw, bins=50)\n",
        "plt.xlabel('Cosine similarity')\n",
        "plt.ylabel('Count')\n",
        "plt.title('Histogram of cosine similarities')\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 472
        },
        "id": "TTjmkjsVg99s",
        "outputId": "8fe030cd-0d83-4571-cb86-dbae9976f86b"
      },
      "execution_count": 27,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Create emply list for genuine and imposter scores\n",
        "genuine_scores = []\n",
        "imposter_scores = []\n",
        "\n",
        "# Populate the lists based on true labels provided\n",
        "for i in range(cosine_similarities_independent.shape[0]):\n",
        "    if cosine_similarities_independent[i][1] == 1:\n",
        "        genuine_scores.append(cosine_similarities_independent[i][0])\n",
        "    else:\n",
        "        imposter_scores.append(cosine_similarities_independent[i][0])"
      ],
      "metadata": {
        "id": "FjrywbUviEyD"
      },
      "execution_count": 28,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Plot histogram for genuine_scores\n",
        "plt.hist(genuine_scores, bins=50, label='Genuine Scores')\n",
        "plt.title('Histogram of Genuine Scores')\n",
        "plt.xlabel('Cosine Similarity')\n",
        "plt.ylabel('Frequency')\n",
        "plt.legend(loc='upper right')\n",
        "plt.show()\n",
        "\n",
        "# Plot histogram for imposter_scores\n",
        "plt.hist(imposter_scores, bins=50, label='Imposter Scores')\n",
        "plt.title('Histogram of Imposter Scores')\n",
        "plt.xlabel('Cosine Similarity')\n",
        "plt.ylabel('Frequency')\n",
        "plt.legend(loc='upper right')\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 927
        },
        "id": "AzxGh-4FjJEj",
        "outputId": "0d6c3015-89a8-4a3b-c51b-3b2303bd390a"
      },
      "execution_count": 30,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAjIAAAHHCAYAAACle7JuAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAABEe0lEQVR4nO3deZxP9f////trhtmYxTLMDJMxY+xRiGTL0mcsbxHZ0xCpjITUm1RIGVGiqN5d3jWDilLS+23LTgvJ/rYkxth3MpuMMfP8/eE7r18vsxqzHW7Xy+V1yeuc53mex3mdeTX3eZ7NZowxAgAAsCCnoi4AAAAgrwgyAADAsggyAADAsggyAADAsggyAADAsggyAADAsggyAADAsggyAADAsggyAADAsggyQCaCgoI0YMCAoi7jjjdt2jQFBwfL2dlZ9913X1GXk28mTJggm81W1GUAdwWCDO540dHRstls2rp1a6bzH374YdWtW/e217Ns2TJNmDDhtvu5W6xcuVIvv/yymjVrpqioKE2ePDnHZX788Uf17NlTlSpVkouLi7y9vdWkSRO98cYbOnv2bCFUXfxdu3ZNM2fO1P333y8vLy/5+PioTp06GjJkiH7//feiLg/IdyWKugCgODpw4ICcnG4t5y9btkyzZ88mzOTS2rVr5eTkpE8//VQuLi45tn/99dc1adIkBQcHa8CAAQoODtbVq1e1bds2vfvuu5ozZ45iYmIKofKcvfrqqxozZkyRrLt79+5avny5+vTpo6efflopKSn6/ffftWTJEj300EOqWbNmkdQFFBSCDJAJV1fXoi7hliUlJalUqVJFXUaunTt3Tu7u7rkKMV999ZUmTZqknj17at68eRmWee+99/Tee+8VVKm3rESJEipRovD/9/rbb79pyZIleuutt/TKK684zJs1a5YuX75caLVcvXpVLi4ut/wHAXCr+AkDMnHzOTIpKSmaOHGiQkND5ebmpnLlyql58+ZatWqVJGnAgAGaPXu2JMlms9lf6ZKSkvTiiy8qMDBQrq6uqlGjht555x3d/PD5v/76S8OHD1f58uXl6empRx99VCdPnpTNZnMY6Uk/B2Pfvn3q27evypQpo+bNm0uSdu/ebR+xcHNzk5+fn5566ildvHjRYV3pffzxxx964okn5O3tLV9fX7322msyxuj48ePq0qWLvLy85Ofnp3fffTdXn93169c1adIkhYSEyNXVVUFBQXrllVeUnJxsb2Oz2RQVFaWkpCT7ZxUdHZ1ln6+//rrKly+f5eiNt7d3piNhy5cvV4sWLVSqVCl5enqqU6dO2rt3r0ObAQMGqHTp0jp58qS6du2q0qVLy9fXV6NHj1Zqaqq93fr162Wz2bR+/XqH5Y8cOZKh/szOkbHZbBo2bJgWL16sunXrytXVVXXq1NGKFSsy1H3y5Ek99dRTqlixor3dZ599luXnky59RKpZs2YZ5jk7O6tcuXIZ1jNo0CAFBATI1dVVVatW1XPPPadr167Z2xw+fFg9evRQ2bJl5eHhoQcffFBLly516Cf9s1mwYIFeffVVVapUSR4eHoqPj5ck/frrr2rfvr28vb3l4eGhVq1a6eeff3boIyEhQSNGjFBQUJBcXV1VoUIFPfLII9q+fXuO2427GyMyuGvExcXpwoULGaanpKTkuOyECRMUGRmpwYMHq3HjxoqPj9fWrVu1fft2PfLII3rmmWd06tQprVq1SvPmzXNY1hijRx99VOvWrdOgQYN033336YcfftBLL72kkydPOowkDBgwQF9//bX69++vBx98UBs2bFCnTp2yrKtHjx4KDQ3V5MmT7aFo1apVOnz4sAYOHCg/Pz/t3btXn3zyifbu3avNmzdn+AXbq1cv1apVS1OmTNHSpUv15ptvqmzZsvrXv/6lNm3a6O2339YXX3yh0aNH64EHHlDLli2z/awGDx6sOXPm6PHHH9eLL76oX3/9VZGRkdq/f7++++47SdK8efP0ySefaMuWLfr3v/8tSXrooYcy7e+PP/7QH3/8ocGDB6t06dLZrvvv5s2bp/DwcIWFhentt9/WlStX9NFHH6l58+basWOHgoKC7G1TU1MVFhamJk2a6J133tHq1av17rvvKiQkRM8991yu15mTn376SYsWLdLQoUPl6emp999/X927d9exY8fsIePs2bN68MEH7cHH19dXy5cv16BBgxQfH68RI0Zk2X+VKlUkSV988YWaNWuW7ajQqVOn1LhxY12+fFlDhgxRzZo1dfLkSX3zzTe6cuWKXFxcdPbsWT300EO6cuWKhg8frnLlymnOnDl69NFH9c033+ixxx5z6HPSpElycXHR6NGjlZycLBcXF61du1YdOnRQw4YNNX78eDk5OSkqKkpt2rTRjz/+qMaNG0uSnn32WX3zzTcaNmyYateurYsXL+qnn37S/v371aBBg9v85HFHM8AdLioqykjK9lWnTh2HZapUqWLCw8Pt7+vXr286deqU7XoiIiJMZl+pxYsXG0nmzTffdJj++OOPG5vNZg4dOmSMMWbbtm1GkhkxYoRDuwEDBhhJZvz48fZp48ePN5JMnz59MqzvypUrGabNnz/fSDIbN27M0MeQIUPs065fv24qV65sbDabmTJlin36n3/+adzd3R0+k8zs3LnTSDKDBw92mD569Ggjyaxdu9Y+LTw83JQqVSrb/owx5vvvvzeSzIwZMxymp6WlmfPnzzu8UlJSjDHGJCQkGB8fH/P00087LHPmzBnj7e3tMD08PNxIMm+88YZD2/vvv980bNjQ/n7dunVGklm3bp1Du9jYWCPJREVF2aelf7Z/J8m4uLjY97cxxuzatctIMh988IF92qBBg4y/v7+5cOGCw/K9e/c23t7eme7fv38mrVq1MpJMxYoVTZ8+fczs2bPN0aNHM7R98sknjZOTk/ntt98y7ccYY0aMGGEkmR9//NE+LyEhwVStWtUEBQWZ1NRUh88mODjYob60tDQTGhpqwsLC7H0ac+NntGrVquaRRx6xT/P29jYRERFZbhuQFQ4t4a4xe/ZsrVq1KsOrXr16OS7r4+OjvXv36uDBg7e83mXLlsnZ2VnDhw93mP7iiy/KGKPly5dLkv0Qw9ChQx3aPf/881n2/eyzz2aY5u7ubv/31atXdeHCBT344IOSlOkw/eDBg+3/dnZ2VqNGjWSM0aBBg+zTfXx8VKNGDR0+fDjLWqQb2ypJo0aNcpj+4osvSlKGQxK5kX544ubRmLi4OPn6+jq8du7cKenGqNTly5fVp08fXbhwwf5ydnZWkyZNtG7dugzrufmzbNGiRY7be6vatWunkJAQ+/t69erJy8vLvh5jjL799lt17txZxhiH2sPCwhQXF5ftoRabzaYffvhBb775psqUKaP58+crIiJCVapUUa9eveznyKSlpWnx4sXq3LmzGjVqlGk/0o392bhxY/thS+nGfhgyZIiOHDmiffv2OSwXHh7u8PO3c+dOHTx4UH379tXFixft25KUlKS2bdtq48aNSktLk3TjZ+zXX3/VqVOnbvFTxd2OQ0u4azRu3DjT/2mXKVMm00NOf/fGG2+oS5cuql69uurWrav27durf//+uQpBR48eVUBAgDw9PR2m16pVyz4//b9OTk6qWrWqQ7tq1apl2ffNbSXp0qVLmjhxohYsWKBz5845zIuLi8vQ/p577nF47+3tLTc3N5UvXz7D9JvPs7lZ+jbcXLOfn598fHzs23or0j+3xMREh+mlS5e2n6O0cuVKTZs2zT4vPXC2adMm0z69vLwc3ru5ucnX19dhWpkyZfTnn3/ecr3Zufmzvnk958+f1+XLl/XJJ5/ok08+ybSPm/fpzVxdXTVu3DiNGzdOp0+f1oYNGzRz5kx9/fXXKlmypD7//HOdP39e8fHxOd524OjRo2rSpEmG6X//2f17Hzf/PKbvh/Dw8CzXERcXpzJlymjq1KkKDw9XYGCgGjZsqI4dO+rJJ59UcHBwtjUCBBkgF1q2bKmYmBh9//33Wrlypf7973/rvffe08cff+wwolHY/v7Xb7qePXvql19+0UsvvaT77rtPpUuXVlpamtq3b2//6/fvnJ2dczVNUoaTk7OSnzeDS79ceM+ePQ7TS5QooXbt2kmSTpw44TAvfTvnzZsnPz+/DH3efO5IVtv7d1lt099PCM5JTp9ret1PPPFElr/8cxOe0/n7+6t3797q3r276tSpo6+//jrbk6pv180/j+nbM23atCxveJg+0tazZ0+1aNFC3333nT2Yvv3221q0aJE6dOhQYDXD+ggyQC6VLVtWAwcO1MCBA5WYmKiWLVtqwoQJ9iCT1S+6KlWqaPXq1UpISHAYlUm/OVn6CZpVqlRRWlqaYmNjFRoaam936NChXNf4559/as2aNZo4caJef/11+/S8HBLLi/RtOHjwoP2vdunGCayXL1+2b+utqFGjhkJDQ7V48WLNmDEjV5eYpx++qVChgj3s3K4yZcpIUoZLmPMyypQVX19feXp6KjU1Nd/qlqSSJUuqXr16OnjwoC5cuKAKFSrIy8srQzi8WZUqVXTgwIEM02/+2c1K+n7w8vLK1fb4+/tr6NChGjp0qM6dO6cGDRrorbfeIsggW5wjA+TCzYdUSpcurWrVqjlcUpz+C/bmX3QdO3ZUamqqZs2a5TD9vffek81ms/9POiwsTJL04YcfOrT74IMPcl1n+l/8N4+czJgxI9d93I6OHTtmur7p06dLUrZXYGVnwoQJunDhgv0Gbze7eXvDwsLk5eWlyZMnZ9r+/Pnzt1xDlSpV5OzsrI0bNzpMv3l/3Q5nZ2d1795d3377baYhI6e6Dx48qGPHjmWYfvnyZW3atEllypSRr6+vnJyc1LVrV/33v//N9I7X6Z9nx44dtWXLFm3atMk+LykpSZ988omCgoJUu3btbOtp2LChQkJC9M4772Q4NPj37UlNTc1w2LNChQoKCAhw+I4BmWFEBsiF2rVr6+GHH1bDhg1VtmxZbd261X6paLqGDRtKkoYPH66wsDA5Ozurd+/e6ty5s1q3bq1x48bpyJEjql+/vlauXKnvv/9eI0aMsP/V2rBhQ3Xv3l0zZszQxYsX7Zdf//HHH5Jyd7jGy8tLLVu21NSpU5WSkqJKlSpp5cqVio2NLYBPJaP69esrPDxcn3zyiS5fvqxWrVppy5YtmjNnjrp27arWrVvnqd++fftqz549ioyM1JYtW9S7d29VrVpVSUlJ2rNnj+bPny9PT0/7qImXl5c++ugj9e/fXw0aNFDv3r3l6+urY8eOaenSpWrWrFmGYJkTb29v9ejRQx988IFsNptCQkK0ZMmSHM9ZuVVTpkzRunXr1KRJEz399NOqXbu2Ll26pO3bt2v16tW6dOlSlsvu2rVLffv2VYcOHdSiRQuVLVtWJ0+e1Jw5c3Tq1CnNmDHDHnYnT56slStXqlWrVhoyZIhq1aql06dPa+HChfrpp5/k4+OjMWPGaP78+erQoYOGDx+usmXLas6cOYqNjdW3336b483unJyc9O9//1sdOnRQnTp1NHDgQFWqVEknT57UunXr5OXlpf/+979KSEhQ5cqV9fjjj6t+/foqXbq0Vq9erd9++y3X9y/CXazIrpcCCkn65deZXWZqjDGtWrXK8fLrN9980zRu3Nj4+PgYd3d3U7NmTfPWW2+Za9eu2dtcv37dPP/888bX19fYbDaHy28TEhLMyJEjTUBAgClZsqQJDQ0106ZNc7gk1RhjkpKSTEREhClbtqwpXbq06dq1qzlw4ICR5HA5dPrlvefPn8+wPSdOnDCPPfaY8fHxMd7e3qZHjx7m1KlTWV7CfXMfWV0WndnnlJmUlBQzceJEU7VqVVOyZEkTGBhoxo4da65evZqr9WRn/fr15vHHHzf+/v6mZMmSxsvLyzRq1MiMHz/enD59OkP7devWmbCwMOPt7W3c3NxMSEiIGTBggNm6dWuOdWR2CfX58+dN9+7djYeHhylTpox55plnzJ49e3J9+XVmlxff/LNmjDFnz541ERERJjAw0JQsWdL4+fmZtm3bmk8++STbz+fs2bNmypQpplWrVsbf39+UKFHClClTxrRp08Z88803GdofPXrUPPnkk8bX19e4urqa4OBgExERYZKTk+1tYmJizOOPP258fHyMm5ubady4sVmyZIlDP+mXXy9cuDDTunbs2GG6detmypUrZ1xdXU2VKlVMz549zZo1a4wxxiQnJ5uXXnrJ1K9f33h6eppSpUqZ+vXrmw8//DDb7QWMMcZmTC7P3gNQJHbu3Kn7779fn3/+ufr161fU5QBAscI5MkAx8tdff2WYNmPGDDk5OeV4R10AuBtxjgxQjEydOlXbtm1T69atVaJECS1fvlzLly/XkCFDFBgYWNTlAUCxw6EloBhZtWqVJk6cqH379ikxMVH33HOP+vfvr3HjxhXJ05QBoLgjyAAAAMviHBkAAGBZBBkAAGBZRXrQPTIyUosWLdLvv/8ud3d3PfTQQ3r77bdVo0YNe5uHH35YGzZscFjumWee0ccff5yrdaSlpenUqVPy9PTM1+e/AACAgmOMUUJCggICArK9+WKRniPTvn179e7dWw888ICuX7+uV155RXv27NG+ffvst3t/+OGHVb16db3xxhv25Tw8PDI8vTYrJ06c4GoPAAAs6vjx46pcuXKW84t0RGbFihUO76Ojo1WhQgVt27bN4Z4ZHh4emT7BNjfSH9J3/PjxXIcfAABQtOLj4xUYGOjwsN3MFKvrOdMfGla2bFmH6V988YU+//xz+fn5qXPnznrttdfk4eGRaR/JyckODxlLSEiQdOPZKwQZAACsJafTQopNkElLS9OIESPUrFkz1a1b1z69b9++qlKligICArR7927985//1IEDB7Ro0aJM+4mMjNTEiRMLq2wAAFCEis19ZJ577jktX75cP/30U7bHwtauXau2bdvq0KFD9qcG/93NIzLpQ1NxcXGMyAAAYBHx8fHy9vbO8fd3sRiRGTZsmJYsWaKNGzdmG2IkqUmTJpKUZZBxdXWVq6trgdQJAACKlyINMsYYPf/88/ruu++0fv16Va1aNcdldu7cKUny9/cv4OoA4O6Tlpama9euFXUZuAuULFlSzs7Ot91PkQaZiIgIffnll/r+++/l6empM2fOSJK8vb3l7u6umJgYffnll+rYsaPKlSun3bt3a+TIkWrZsqXq1atXlKUDwB3n2rVrio2NVVpaWlGXgruEj4+P/Pz8bus+b0V6jkxWhUdFRWnAgAE6fvy4nnjiCe3Zs0dJSUkKDAzUY489pldffTXX57vk9hgbANzNjDE6duyYUlJScrwBGXC7jDG6cuWKzp07Jx8fn0yPsljiHJmcMlRgYGCGu/oCAPLf9evXdeXKFQUEBGR5ewsgP7m7u0uSzp07pwoVKuT5MBORGwCg1NRUSZKLi0sRV4K7SXpoTklJyXMfBBkAgB3PpENhyo+fN4IMAACwLIIMAAD5yGazafHixUVdxl2jWNwQDwBQPAWNWVqo6zsypdMtL3PmzBlFRkZq6dKlOnHihLy9vVWtWjU98cQTCg8PL/STl0+fPq0yZcoU6DpSU1M1bdo0RUdH6+jRo3J3d1doaKiefvppDR48uEDXXdwQZAAAlnX48GE1a9ZMPj4+mjx5su699165urrqf//7nz755BNVqlRJjz76aKHW5OfnV+DrmDhxov71r39p1qxZatSokeLj47V161b9+eefBbbOa9euFcuTwTm0BACwrKFDh6pEiRLaunWrevbsqVq1aik4OFhdunTR0qVL1blzZ3vby5cva/DgwfL19ZWXl5fatGmjXbt22edPmDBB9913n+bNm6egoCB5e3urd+/eSkhIsLcJCgrSjBkzHGq47777NGHCBPv7vx9aOnLkiGw2mxYtWqTWrVvLw8ND9evX16ZNmxz6+Omnn9SiRQu5u7srMDBQw4cPV1JSUpbb/Z///EdDhw5Vjx49VLVqVdWvX1+DBg3S6NGj7W3S0tI0depUVatWTa6urrrnnnv01ltv2ef/73//U5s2beTu7q5y5cppyJAhSkxMtM8fMGCAunbtqrfeeksBAQGqUaOGJOn48ePq2bOnfHx8VLZsWXXp0kVHjhyxL7d+/Xo1btxYpUqVko+Pj5o1a6ajR49muS23iyADALCkixcvauXKlYqIiFCpUqUybfP3q2J69Oihc+fOafny5dq2bZsaNGigtm3b6tKlS/Y2MTExWrx4sZYsWaIlS5Zow4YNmjJlym3XOm7cOI0ePVo7d+5U9erV1adPH12/ft2+zvbt26t79+7avXu3vvrqK/30008aNmxYlv35+flp7dq1On/+fJZtxo4dqylTpui1117Tvn379OWXX6pixYqSpKSkJIWFhalMmTL67bfftHDhQq1evTrDOtesWaMDBw5o1apVWrJkiVJSUhQWFiZPT0/9+OOP+vnnn1W6dGm1b99e165d0/Xr19W1a1e1atVKu3fv1qZNmzRkyJACvRqOIAMAsKRDhw7JGGMfKUhXvnx5lS5dWqVLl9Y///lPSTdGPLZs2aKFCxeqUaNGCg0N1TvvvCMfHx9988039mXT0tIUHR2tunXrqkWLFurfv7/WrFlz27WOHj1anTp1UvXq1TVx4kQdPXpUhw4dkiRFRkaqX79+GjFihEJDQ/XQQw/p/fff19y5c3X16tVM+5s+fbrOnz8vPz8/1atXT88++6yWL19un5+QkKCZM2dq6tSpCg8PV0hIiJo3b24/f+bLL7/U1atXNXfuXNWtW1dt2rTRrFmzNG/ePJ09e9bej6u7h16Y8I5SvSsp1buSpn30mf66dl3DJ7wjUyZQKZ7+ioqK0rFjx7R+/XrFx8crLi5O//jHPxQSEqJatWopPDxc99xzz21/hlkhyAAA7ihbtmzRzp07VadOHSUnJ0uSdu3apcTERJUrV84eckqXLq3Y2FjFxMTYlw0KCpKnp6f9vb+/v86dO3fbNf39+YDpt+NP73fXrl2Kjo52qCssLExpaWmKjY3NtL/atWtrz5492rx5s5566imdO3dOnTt3tgeV/fv3Kzk5WW3bts10+f3796t+/foOI1nNmjVTWlqaDhw4YJ8WWrO2Sv7tvJg/9u3R8SOH1bRmoB6sUVkP1qissmXL6urVq4qJiVHZsmU1YMAAhYWFqXPnzpo5c6ZOnz6dx08tdzjZFwBgSdWqVZPNZnP4xStJwcHBkv7/W+BLUmJiovz9/bV+/foM/fj4+Nj/XbJkSYd5NpvN4SGaTk5OGR6vk5u70v693/TDLOn9JiYm6plnntHw4cMzLJfdSIaTk5MeeOABPfDAAxoxYoQ+//xz9e/fX+PGjXPY9tvh7u54xdeVpCTVuvc+Rb7/iX1aTf8bz0Hy9fWVdON5icOHD9eKFSv01Vdf6dVXX9WqVav04IMP5ktNNyPIAAAsqVy5cnrkkUc0a9YsPf/881meJyNJDRo00JkzZ1SiRAkFBQXleZ2+vr4OIwzx8fFZjprkVoMGDbRv3z5Vq1bttvqpXbu2pBvnv4SGhsrd3V1r1qzJ9HLsWrVqKTo6WklJSfbP7eeff5aTk1OGQ3UOy91bXz/89zuVLV9epT1vBJhqlX0ytLv//vt1//33a+zYsWratKm+/PLLAgsyHFoCAFjWhx9+qOvXr6tRo0b66quvtH//fh04cECff/65fv/9d/uDCNu1a6emTZuqa9euWrlypY4cOaJffvlF48aN09atW3O9vjZt2mjevHn68ccf9b///U/h4eF5fthhun/+85/65ZdfNGzYMO3cuVMHDx7U999/n+3Jvo8//rjee+89/frrrzp69KjWr1+viIgIVa9eXTVr1pSbm5v++c9/6uWXX9bcuXMVExOjzZs369NPP5Uk9evXT25ubgoPD9eePXu0bt06Pf/88+rfv7/9hODMdHysh3zKltMLg/pp+6+/6MSxG+sePny4Tpw4odjYWI0dO1abNm3S0aNHtXLlSh08eFC1atW6rc8oO4zIAAAsKyQkRDt27NDkyZM1duxYnThxQq6urqpdu7ZGjx6toUOHSrpxOGfZsmUaN26cBg4caD9RtmXLltn+4r7Z2LFjFRsbq3/84x/y9vbWpEmTbntEpl69etqwYYPGjRunFi1ayBijkJAQ9erVK8tlwsLCNH/+fEVGRiouLk5+fn5q06aNJkyYoBIlbvxqf+2111SiRAm9/vrrOnXqlPz9/fXss89KuvGwxh9++EEvvPCCHnjgAXl4eKh79+6aPn16trW6u3so6pulmhE5QaOGPKmkpERVrlRJbdu2lZeXl/766y/9/vvvmjNnji5evCh/f39FRETomWeeua3PKDs2c/PBvjtMfHy8vL29FRcXJy8vr6IuBwCKpatXryo2NlZVq1aVm5tbUZeDYmL3ics5tqmXyaGl3Mru5y63v785tAQAACyLIAMAACyLIAMAACyLIAMAACyLIAMAsLvDr/9AMZMfP28EGQCA/V4o165dK+JKcDe5cuWKpIx3VL4V3EcGAKASJUrIw8ND58+fV8mSJeXkxN+5kMz1nINtVg+2zLZfY3TlyhWdO3dOPj4+t3VTQYIMAEA2m03+/v6KjY3V0aNHi7ocFBPn/vwrxzYuf+X9uU4+Pj7y8/PL8/ISQQYA8P+4uLgoNDSUw0uwG7xofY5t1rz4cJ76Llmy5G0/3kEiyAAA/sbJyYk7+8LuZEJqjm2K+ueFg6AAAMCyCDIAAMCyCDIAAMCyCDIAAMCyCDIAAMCyCDIAAMCyCDIAAMCyCDIAAMCyCDIAAMCyCDIAAMCyCDIAAMCyCDIAAMCyCDIAAMCyCDIAAMCyCDIAAMCyCDIAAMCyCDIAAMCyCDIAAMCyCDIAAMCyCDIAAMCyCDIAAMCyCDIAAMCyCDIAAMCyCDIAAMCyCDIAAMCyCDIAAMCyCDIAAMCyCDIAAMCyCDIAAMCyCDIAAMCyCDIAAMCyCDIAAMCyCDIAAMCyCDIAAMCyCDIAAMCyCDIAAMCyCDIAAMCyijTIREZG6oEHHpCnp6cqVKigrl276sCBAw5trl69qoiICJUrV06lS5dW9+7ddfbs2SKqGAAAFCdFGmQ2bNigiIgIbd68WatWrVJKSor+7//+T0lJSfY2I0eO1H//+18tXLhQGzZs0KlTp9StW7cirBoAABQXJYpy5StWrHB4Hx0drQoVKmjbtm1q2bKl4uLi9Omnn+rLL79UmzZtJElRUVGqVauWNm/erAcffLAoygYAAMVEsTpHJi4uTpJUtmxZSdK2bduUkpKidu3a2dvUrFlT99xzjzZt2pRpH8nJyYqPj3d4AQCAO1OxCTJpaWkaMWKEmjVrprp160qSzpw5IxcXF/n4+Di0rVixos6cOZNpP5GRkfL29ra/AgMDC7p0AABQRIpNkImIiNCePXu0YMGC2+pn7NixiouLs7+OHz+eTxUCAIDipkjPkUk3bNgwLVmyRBs3blTlypXt0/38/HTt2jVdvnzZYVTm7Nmz8vPzy7QvV1dXubq6FnTJAACgGCjSERljjIYNG6bvvvtOa9euVdWqVR3mN2zYUCVLltSaNWvs0w4cOKBjx46padOmhV0uAAAoZop0RCYiIkJffvmlvv/+e3l6etrPe/H29pa7u7u8vb01aNAgjRo1SmXLlpWXl5eef/55NW3alCuWAABA0QaZjz76SJL08MMPO0yPiorSgAEDJEnvvfeenJyc1L17dyUnJyssLEwffvhhIVcKAACKoyINMsaYHNu4ublp9uzZmj17diFUBAAArKTYXLUEAABwqwgyAADAsggyAADAsggyAADAsggyAADAsggyAADAsggyAADAsggyAADAsorFQyMBAEDhChqztKhLyBeMyAAAAMsiyAAAAMsiyAAAAMsiyAAAAMsiyAAAAMsiyAAAAMsiyAAAAMsiyAAAAMsiyAAAAMsiyAAAAMsiyAAAAMsiyAAAAMsiyAAAAMsiyAAAAMsiyAAAAMsiyAAAAMsiyAAAAMsiyAAAAMsiyAAAAMsiyAAAAMsiyAAAAMsiyAAAAMsiyAAAAMsiyAAAAMsiyAAAAMsiyAAAAMsiyAAAAMsiyAAAAMsqUdQFAACA/BU0ZmlRl1BoGJEBAACWRZABAACWRZABAACWRZABAACWRZABAACWRZABAACWRZABAACWRZABAACWRZABAACWRZABAACWRZABAACWRZABAACWRZABAACWRZABAACWRZABAACWRZABAACWRZABAACWRZABAACWRZABAACWRZABAACWRZABAACWRZABAACWRZABAACWRZABAACWVaRBZuPGjercubMCAgJks9m0ePFih/kDBgyQzWZzeLVv375oigUAAMVOkQaZpKQk1a9fX7Nnz86yTfv27XX69Gn7a/78+YVYIQAAKM5KFOXKO3TooA4dOmTbxtXVVX5+foVUEQAAsJJif47M+vXrVaFCBdWoUUPPPfecLl68WNQlAQCAYqJIR2Ry0r59e3Xr1k1Vq1ZVTEyMXnnlFXXo0EGbNm2Ss7NzpsskJycrOTnZ/j4+Pr6wygUAAIWsWAeZ3r172/997733ql69egoJCdH69evVtm3bTJeJjIzUxIkTC6tEAABQhIr9oaW/Cw4OVvny5XXo0KEs24wdO1ZxcXH21/HjxwuxQgAAUJjyNCJz+PBhBQcH53ctOTpx4oQuXrwof3//LNu4urrK1dW1EKsCAABFJU8jMtWqVVPr1q31+eef6+rVq3leeWJionbu3KmdO3dKkmJjY7Vz504dO3ZMiYmJeumll7R582YdOXJEa9asUZcuXVStWjWFhYXleZ0AAODOkacgs337dtWrV0+jRo2Sn5+fnnnmGW3ZsuWW+9m6davuv/9+3X///ZKkUaNG6f7779frr78uZ2dn7d69W48++qiqV6+uQYMGqWHDhvrxxx8ZcQEAAJIkmzHG5HXh69ev6z//+Y+io6O1YsUKVa9eXU899ZT69+8vX1/f/Kwzz+Lj4+Xt7a24uDh5eXkVdTkAABS4oDFLC21dR6Z0KpB+c/v7+7ZO9i1RooS6deumhQsX6u2339ahQ4c0evRoBQYG6sknn9Tp06dvp3sAAIBs3VaQ2bp1q4YOHSp/f39Nnz5do0ePVkxMjFatWqVTp06pS5cu+VUnAABABnm6amn69OmKiorSgQMH1LFjR82dO1cdO3aUk9ONXFS1alVFR0crKCgoP2sFAABwkKcg89FHH+mpp57SgAEDsrwUukKFCvr0009vqzgAAIDs5CnIHDx4MMc2Li4uCg8Pz0v3AAAAuZKnc2SioqK0cOHCDNMXLlyoOXPm3HZRAAAAuZGnIBMZGany5ctnmF6hQgVNnjz5tosCAADIjTwFmWPHjqlq1aoZplepUkXHjh277aIAAAByI09BpkKFCtq9e3eG6bt27VK5cuVuuygAAIDcyFOQ6dOnj4YPH65169YpNTVVqampWrt2rV544QX17t07v2sEAADIVJ6uWpo0aZKOHDmitm3bqkSJG12kpaXpySef5BwZAABQaPIUZFxcXPTVV19p0qRJ2rVrl9zd3XXvvfeqSpUq+V0fAABAlvIUZNJVr15d1atXz69aAAAAbkmegkxqaqqio6O1Zs0anTt3TmlpaQ7z165dmy/FAQAAZCdPQeaFF15QdHS0OnXqpLp168pms+V3XQAAADnKU5BZsGCBvv76a3Xs2DG/6wEAAMi1PF1+7eLiomrVquV3LQAAALckT0HmxRdf1MyZM2WMye96AAAAci1Ph5Z++uknrVu3TsuXL1edOnVUsmRJh/mLFi3Kl+IAAACyk6cg4+Pjo8ceeyy/awEAALgleQoyUVFR+V0HAADALcvTOTKSdP36da1evVr/+te/lJCQIEk6deqUEhMT8604AACA7ORpRObo0aNq3769jh07puTkZD3yyCPy9PTU22+/reTkZH388cf5XScAAEAGeRqReeGFF9SoUSP9+eefcnd3t09/7LHHtGbNmnwrDgAAIDt5GpH58ccf9csvv8jFxcVhelBQkE6ePJkvhQEAAOQkTyMyaWlpSk1NzTD9xIkT8vT0vO2iAAAAciNPQeb//u//NGPGDPt7m82mxMREjR8/nscWAACAQpOnQ0vvvvuuwsLCVLt2bV29elV9+/bVwYMHVb58ec2fPz+/awQAAMhUnoJM5cqVtWvXLi1YsEC7d+9WYmKiBg0apH79+jmc/AsAAFCQ8hRkJKlEiRJ64okn8rMWAACAW5KnIDN37txs5z/55JN5KgYAAOBW5CnIvPDCCw7vU1JSdOXKFbm4uMjDw4MgAwAACkWerlr6888/HV6JiYk6cOCAmjdvzsm+AACg0OT5WUs3Cw0N1ZQpUzKM1gAAABSUfAsy0o0TgE+dOpWfXQIAAGQpT+fI/Oc//3F4b4zR6dOnNWvWLDVr1ixfCgMAAMhJnoJM165dHd7bbDb5+vqqTZs2evfdd/OjLgAAgBzlKcikpaXldx0AAAC3LM83xAMAAIUvaMzSoi6hWMlTkBk1alSu206fPj0vqwAAAMhRnoLMjh07tGPHDqWkpKhGjRqSpD/++EPOzs5q0KCBvZ3NZsufKgEAADKRpyDTuXNneXp6as6cOSpTpoykGzfJGzhwoFq0aKEXX3wxX4sEAADITJ7uI/Puu+8qMjLSHmIkqUyZMnrzzTe5agkAABSaPAWZ+Ph4nT9/PsP08+fPKyEh4baLAgAAyI08BZnHHntMAwcO1KJFi3TixAmdOHFC3377rQYNGqRu3brld40AAACZytM5Mh9//LFGjx6tvn37KiUl5UZHJUpo0KBBmjZtWr4WCAAAkJU8BRkPDw99+OGHmjZtmmJiYiRJISEhKlWqVL4WBwAAkJ3bemjk6dOndfr0aYWGhqpUqVIyxuRXXQAAADnKU5C5ePGi2rZtq+rVq6tjx446ffq0JGnQoEFceg0AAApNnoLMyJEjVbJkSR07dkweHh726b169dKKFSvyrTgAAIDs5OkcmZUrV+qHH35Q5cqVHaaHhobq6NGj+VIYAABATvI0IpOUlOQwEpPu0qVLcnV1ve2iAAAAciNPQaZFixaaO3eu/b3NZlNaWpqmTp2q1q1b51txAAAA2cnToaWpU6eqbdu22rp1q65du6aXX35Ze/fu1aVLl/Tzzz/nd40AAACZytOITN26dfXHH3+oefPm6tKli5KSktStWzft2LFDISEh+V0jAABApm55RCYlJUXt27fXxx9/rHHjxhVETQAAALlyyyMyJUuW1O7duwuiFgAAgFuSp0NLTzzxhD799NP8rgUAAOCW5Olk3+vXr+uzzz7T6tWr1bBhwwzPWJo+fXq+FAcAAJCdWwoyhw8fVlBQkPbs2aMGDRpIkv744w+HNjabLf+qAwAAyMYtBZnQ0FCdPn1a69atk3TjkQTvv/++KlasWCDFAQAAZOeWzpG5+enWy5cvV1JSUr4WBAAAkFt5Otk33c3BBgAAoDDdUpCx2WwZzoG5nXNiNm7cqM6dOysgIEA2m02LFy92mG+M0euvvy5/f3+5u7urXbt2OnjwYJ7XBwAA7iy3dI6MMUYDBgywPxjy6tWrevbZZzNctbRo0aJc9ZeUlKT69evrqaeeUrdu3TLMnzp1qt5//33NmTNHVatW1WuvvaawsDDt27dPbm5ut1I6AAC4A91SkAkPD3d4/8QTT9zWyjt06KAOHTpkOs8YoxkzZujVV19Vly5dJElz585VxYoVtXjxYvXu3fu21g0AAKzvloJMVFRUQdWRQWxsrM6cOaN27drZp3l7e6tJkybatGlTlkEmOTlZycnJ9vfx8fEFXisAACgaebohXmE4c+aMJGW4tLtixYr2eZmJjIzUxIkTC7Q2AAAKQtCYpUVdguXc1lVLxdHYsWMVFxdnfx0/fryoSwIAAAWk2AYZPz8/SdLZs2cdpp89e9Y+LzOurq7y8vJyeAEAgDtTsQ0yVatWlZ+fn9asWWOfFh8fr19//VVNmzYtwsoAAEBxUaTnyCQmJurQoUP297Gxsdq5c6fKli2re+65RyNGjNCbb76p0NBQ++XXAQEB6tq1a9EVDQAAio0iDTJbt25V69at7e9HjRol6cZl3tHR0Xr55ZeVlJSkIUOG6PLly2revLlWrFjBPWQAAIAkyWbu8OcMxMfHy9vbW3FxcZwvAwAo1qx41dKRKZ0KpN/c/v4utufIAAAA5IQgAwAALIsgAwAALIsgAwAALIsgAwAALIsgAwAALIsgAwAALIsgAwAALIsgAwAALIsgAwAALIsgAwAALIsgAwAALIsgAwAALIsgAwAALIsgAwAALIsgAwAALIsgAwAALIsgAwAALIsgAwAALIsgAwAALIsgAwAALIsgAwAALKtEURcAAMDdIGjM0qIu4Y7EiAwAALAsggwAALAsggwAALAsggwAALAsggwAALAsggwAALAsggwAALAsggwAALAsggwAALAsggwAALAsggwAALAsggwAALAsggwAALAsggwAALCsEkVdAAAAxVnQmKU5tjkypVMhVILMMCIDAAAsiyADAAAsiyADAAAsiyADAAAsiyADAAAsiyADAAAsiyADAAAsiyADAAAsiyADAAAsiyADAAAsiyADAAAsiyADAAAsiyADAAAsiyADAAAsq0RRFwAAQEEIGrM0xzZHpnQqtHWhYDAiAwAALIsgAwAALIsgAwAALIsgAwAALIsgAwAALIsgAwAALIsgAwAALIsgAwAALIsgAwAALKtYB5kJEybIZrM5vGrWrFnUZQEAgGKi2D+ioE6dOlq9erX9fYkSxb5kAABQSIp9KihRooT8/PyKugwAAFAMFetDS5J08OBBBQQEKDg4WP369dOxY8eybZ+cnKz4+HiHFwAAuDMV6yDTpEkTRUdHa8WKFfroo48UGxurFi1aKCEhIctlIiMj5e3tbX8FBgYWYsUAAKAwFesg06FDB/Xo0UP16tVTWFiYli1bpsuXL+vrr7/OcpmxY8cqLi7O/jp+/HghVgwAAApTsT9H5u98fHxUvXp1HTp0KMs2rq6ucnV1LcSqAABAUSnWIzI3S0xMVExMjPz9/Yu6FAAAUAwU6yAzevRobdiwQUeOHNEvv/yixx57TM7OzurTp09RlwYAAIqBYn1o6cSJE+rTp48uXrwoX19fNW/eXJs3b5avr29RlwYAAIqBYh1kFixYUNQlAACAYqxYH1oCAADIDkEGAABYFkEGAABYFkEGAABYFkEGAABYFkEGAABYFkEGAABYFkEGAABYVrG+IR4A4M4SNGZpjm2OTOmUL/3kRn71g6LDiAwAALAsggwAALAsggwAALAsggwAALAsggwAALAsggwAALAsggwAALAsggwAALAsggwAALAsggwAALAsggwAALAsggwAALAsggwAALAsggwAALCsEkVdAACg+AsaszTHNkemdCq0dQHpGJEBAACWRZABAACWRZABAACWRZABAACWRZABAACWRZABAACWRZABAACWRZABAACWRZABAACWRZABAACWRZABAACWRZABAACWRZABAACWxdOvAUCF+3Tn/JJfT4nmqdWwMkZkAACAZRFkAACAZRFkAACAZRFkAACAZRFkAACAZRFkAACAZRFkAACAZRFkAACAZRFkAACAZRFkAACAZRFkAACAZRFkAACAZRFkAACAZRFkAACAZZUo6gKsLDePrD8ypVMhVAJYT359f3LTD7LHZwgrY0QGAABYFkEGAABYFkEGAABYFkEGAABYFkEGAABYFkEGAABYFkEGAABYFkEGAABYFkEGAABYliWCzOzZsxUUFCQ3Nzc1adJEW7ZsKeqSAABAMVDsg8xXX32lUaNGafz48dq+fbvq16+vsLAwnTt3rqhLAwAARazYB5np06fr6aef1sCBA1W7dm19/PHH8vDw0GeffVbUpQEAgCJWrIPMtWvXtG3bNrVr184+zcnJSe3atdOmTZuKsDIAAFAcFOunX1+4cEGpqamqWLGiw/SKFSvq999/z3SZ5ORkJScn29/HxcVJkuLj4/O9vrTkKzm2KYj1AneC/Pr+5Kaf/FLcvs+Fue1AVgrqe5HerzEm23bFOsjkRWRkpCZOnJhhemBgYBFUI3nPKJLVAneE4vb9KW71AMVBQX8vEhIS5O3tneX8Yh1kypcvL2dnZ509e9Zh+tmzZ+Xn55fpMmPHjtWoUaPs79PS0nTp0iWVK1dONputQOstKPHx8QoMDNTx48fl5eVV1OUgC+wn62BfWQP7yToKYl8ZY5SQkKCAgIBs2xXrIOPi4qKGDRtqzZo16tq1q6QbwWTNmjUaNmxYpsu4urrK1dXVYZqPj08BV1o4vLy8+DJbAPvJOthX1sB+so783lfZjcSkK9ZBRpJGjRql8PBwNWrUSI0bN9aMGTOUlJSkgQMHFnVpAACgiBX7INOrVy+dP39er7/+us6cOaP77rtPK1asyHACMAAAuPsU+yAjScOGDcvyUNLdwNXVVePHj89wyAzFC/vJOthX1sB+so6i3Fc2k9N1TQAAAMVUsb4hHgAAQHYIMgAAwLIIMgAAwLIIMgAAwLIIMkVg9uzZCgoKkpubm5o0aaItW7Zk237GjBmqUaOG3N3dFRgYqJEjR+rq1au31SdyJ7/31YQJE2Sz2RxeNWvWLOjNuOPdyn5KSUnRG2+8oZCQELm5ual+/fpasWLFbfWJ3MvvfcV3Kv9t3LhRnTt3VkBAgGw2mxYvXpzjMuvXr1eDBg3k6uqqatWqKTo6OkObAvtOGRSqBQsWGBcXF/PZZ5+ZvXv3mqefftr4+PiYs2fPZtr+iy++MK6uruaLL74wsbGx5ocffjD+/v5m5MiRee4TuVMQ+2r8+PGmTp065vTp0/bX+fPnC2uT7ki3up9efvllExAQYJYuXWpiYmLMhx9+aNzc3Mz27dvz3CdypyD2Fd+p/Lds2TIzbtw4s2jRIiPJfPfdd9m2P3z4sPHw8DCjRo0y+/btMx988IFxdnY2K1assLcpyO8UQaaQNW7c2ERERNjfp6ammoCAABMZGZlp+4iICNOmTRuHaaNGjTLNmjXLc5/InYLYV+PHjzf169cvkHrvVre6n/z9/c2sWbMcpnXr1s3069cvz30idwpiX/GdKli5CTIvv/yyqVOnjsO0Xr16mbCwMPv7gvxOcWipEF27dk3btm1Tu3bt7NOcnJzUrl07bdq0KdNlHnroIW3bts0+BHf48GEtW7ZMHTt2zHOfyFlB7Kt0Bw8eVEBAgIKDg9WvXz8dO3as4DbkDpeX/ZScnCw3NzeHae7u7vrpp5/y3CdyVhD7Kh3fqaK1adMmh/0qSWFhYfb9WtDfKYJMIbpw4YJSU1MzPF6hYsWKOnPmTKbL9O3bV2+88YaaN2+ukiVLKiQkRA8//LBeeeWVPPeJnBXEvpKkJk2aKDo6WitWrNBHH32k2NhYtWjRQgkJCQW6PXeqvOynsLAwTZ8+XQcPHlRaWppWrVqlRYsW6fTp03nuEzkriH0l8Z0qDs6cOZPpfo2Pj9dff/1V4N8pgkwxt379ek2ePFkffvihtm/frkWLFmnp0qWaNGlSUZeGm+RmX3Xo0EE9evRQvXr1FBYWpmXLluny5cv6+uuvi7Dyu8vMmTMVGhqqmjVrysXFRcOGDdPAgQPl5MT/Doub3OwrvlOwxLOW7hTly5eXs7Ozzp496zD97Nmz8vPzy3SZ1157Tf3799fgwYMlSffee6+SkpI0ZMgQjRs3Lk99ImcFsa8y+0Xp4+Oj6tWr69ChQ/m/EXeBvOwnX19fLV68WFevXtXFixcVEBCgMWPGKDg4OM99ImcFsa8yw3eq8Pn5+WW6X728vOTu7i5nZ+cC/U7xJ0ghcnFxUcOGDbVmzRr7tLS0NK1Zs0ZNmzbNdJkrV65k+AXo7OwsSTLG5KlP5Kwg9lVmEhMTFRMTI39//3yq/O5yOz//bm5uqlSpkq5fv65vv/1WXbp0ue0+kbWC2FeZ4TtV+Jo2beqwXyVp1apV9v1a4N+p2z5dGLdkwYIFxtXV1URHR5t9+/aZIUOGGB8fH3PmzBljjDH9+/c3Y8aMsbcfP3688fT0NPPnzzeHDx82K1euNCEhIaZnz5657hN5UxD76sUXXzTr1683sbGx5ueffzbt2rUz5cuXN+fOnSv07btT3Op+2rx5s/n2229NTEyM2bhxo2nTpo2pWrWq+fPPP3PdJ/KmIPYV36n8l5CQYHbs2GF27NhhJJnp06ebHTt2mKNHjxpjjBkzZozp37+/vX365dcvvfSS2b9/v5k9e3aml18X1HeKIFMEPvjgA3PPPfcYFxcX07hxY7N582b7vFatWpnw8HD7+5SUFDNhwgQTEhJi3NzcTGBgoBk6dKjDFzmnPpF3+b2vevXqZfz9/Y2Li4upVKmS6dWrlzl06FAhbtGd6Vb20/r1602tWrWMq6urKVeunOnfv785efLkLfWJvMvvfcV3Kv+tW7fOSMrwSt834eHhplWrVhmWue+++4yLi4sJDg42UVFRGfotqO+UzZgsxrwBAACKOc6RAQAAlkWQAQAAlkWQAQAAlkWQAQAAlkWQAQAAlkWQAQAAlkWQAQAAlkWQAXBboqOj5ePjU9Rl6MiRI7LZbNq5c+dt9fPwww9rxIgR9vdBQUGaMWPGbfUpSQMGDFDXrl1vux8AjggywB3uzJkzev755xUcHCxXV1cFBgaqc+fOGZ6Nkle9evXSH3/8kS99ZSc2NlZ9+/ZVQECA3NzcVLlyZXXp0kW///67JCkwMFCnT59W3bp1b2s9ixYtKpCny8+cOVPR0dH29zcHJgB5w9OvgTvYkSNH1KxZM/n4+GjatGm69957lZKSoh9++EERERH2EHA73N3d5e7ung/VZi0lJUWPPPKIatSooUWLFsnf318nTpzQ8uXLdfnyZUk3HtCZH0/SLVu27G338Xepqamy2Wzy9vbO134B/D/58qADAMVShw4dTKVKlUxiYmKGeX9/BtTRo0fNo48+akqVKmU8PT1Njx49HB7mtnPnTvPwww+b0qVLG09PT9OgQQPz22+/GWOMiYqKMt7e3va248ePN/Xr1zdz5841VapUMV5eXqZXr14mPj7e3iY1NdVMnjzZBAUFGTc3N1OvXj2zcOHCLLcj/eF1R44cybJNbGyskWR27NhhjPn/nxezYsUKc9999xk3NzfTunVrc/bsWbNs2TJTs2ZN4+npafr06WOSkpLs/bRq1cq88MIL9vdVqlQx7733nv39u+++a+rWrWs8PDxM5cqVzXPPPWcSEhLs89M/j++//97UqlXLODs7m9jYWBMeHm66dOlijLnxrBrd9Bybw4cPm5CQEDNt2rRMt/3gwYNZbjtwN+PQEnCHunTpklasWKGIiAiVKlUqw/z081rS0tLUpUsXXbp0SRs2bNCqVat0+PBh9erVy962X79+qly5sn777Tdt27ZNY8aMUcmSJbNcd0xMjBYvXqwlS5ZoyZIl2rBhg6ZMmWKfHxkZqblz5+rjjz/W3r17NXLkSD3xxBPasGFDpv35+vrKyclJ33zzjVJTU2/pc5gwYYJmzZqlX375RcePH1fPnj01Y8YMffnll1q6dKlWrlypDz74INf9OTk56f3339fevXs1Z84crV27Vi+//LJDmytXrujtt9/Wv//9b+3du1cVKlRwmD9z5kw1bdpUTz/9tE6fPq3Tp0/rnnvu0VNPPaWoqCiHtlFRUWrZsqWqVat2S9sN3DWKOkkBKBi//vqrkWQWLVqUbbuVK1caZ2dnc+zYMfu0vXv3Gklmy5YtxhhjPD09TXR0dKbLZzYi4+Hh4TAC89JLL5kmTZoYY4y5evWq8fDwML/88otDP4MGDTJ9+vTJss5Zs2YZDw8P4+npaVq3bm3eeOMNExMTY5+f1YjM6tWr7W0iIyONJIflnnnmGRMWFmZ/n9OIzM0WLlxoypUr5/B5SDI7d+50aPf3EZnM1mOMMSdPnjTOzs7m119/NcYYc+3aNVO+fPksP3sAjMgAdyyTywfb79+/X4GBgQoMDLRPq127tnx8fLR//35J0qhRozR48GC1a9dOU6ZMUUxMTLZ9BgUFydPT0/7e399f586dkyQdOnRIV65c0SOPPKLSpUvbX3Pnzs2234iICJ05c0ZffPGFmjZtqoULF6pOnTpatWpVtrXUq1fP/u+KFSvKw8NDwcHBDtPSa8uN1atXq23btqpUqZI8PT3Vv39/Xbx4UVeuXLG3cXFxcVhvbgUEBKhTp0767LPPJEn//e9/lZycrB49etxyX8DdgiAD3KFCQ0Nls9ny5YTeCRMmaO/everUqZPWrl2r2rVr67vvvsuy/c2HnWw2m9LS0iRJiYmJkqSlS5dq586d9te+ffv0zTffZFuHp6enOnfurLfeeku7du1SixYt9Oabb2a7zN9rsdls2daWkyNHjugf//iH6tWrp2+//Vbbtm3T7NmzJUnXrl2zt3N3d5fNZstVnzcbPHiwFixYoL/++ktRUVHq1auXPDw88tQXcDcgyAB3qLJlyyosLEyzZ89WUlJShvnpV/vUqlVLx48f1/Hjx+3z9u3bp8uXL6t27dr2adWrV9fIkSO1cuVKdevWLcO5HLlVu3Ztubq66tixY6pWrZrD6++jQjmx2WyqWbNmpttWULZt26a0tDS9++67evDBB1W9enWdOnUqT325uLhker5Px44dVapUKX300UdasWKFnnrqqdstG7ijEWSAO9js2bOVmpqqxo0b69tvv9XBgwe1f/9+vf/++2ratKkkqV27drr33nvVr18/bd++XVu2bNGTTz6pVq1aqVGjRvrrr780bNgwrV+/XkePHtXPP/+s3377TbVq1cpTTZ6enho9erRGjhypOXPmKCYmRtu3b9cHH3ygOXPmZLrMzp071aVLF33zzTfat2+fDh06pE8//VSfffaZunTpkufP51ZVq1ZNKSkp+uCDD3T48GHNmzdPH3/8cZ76CgoK0q+//qojR47owoUL9lEhZ2dnDRgwQGPHjlVoaKh9PwHIHEEGuIMFBwdr+/btat26tV588UXVrVtXjzzyiNasWaOPPvpI0o2Rje+//15lypRRy5Yt1a5dOwUHB+urr76SdOMX68WLF/Xkk0+qevXq6tmzpzp06KCJEyfmua5JkybptddeU2RkpGrVqqX27dtr6dKlqlq1aqbtK1eurKCgIE2cOFFNmjRRgwYNNHPmTE2cOFHjxo3Lcx23qn79+po+fbrefvtt1a1bV1988YUiIyPz1Nfo0aPl7Oys2rVry9fXV8eOHbPPGzRokK5du6aBAwfmV+nAHctmcntGIACgUPz4449q27atjh8/rooVKxZ1OUCxRpABgGIiOTlZ58+fV3h4uPz8/PTFF18UdUlAscehJQAoJubPn68qVaro8uXLmjp1alGXA1gCIzIAAMCyGJEBAACWRZABAACWRZABAACWRZABAACWRZABAACWRZABAACWRZABAACWRZABAACWRZABAACW9f8BTE4Fknb9XukAAAAASUVORK5CYII=\n"
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# create arrays to store true labels and similarity scores\n",
        "true_labels = cosine_similarities_t[1]\n",
        "scores = cosine_similarities_t[0]\n",
        "\n",
        "# create arrays to store FPR and TPR for different thresholds\n",
        "fpr = []\n",
        "tpr = []\n",
        "thresholds = []\n",
        "\n",
        "# calculate the FPR and TPR for different thresholds\n",
        "fpr, tpr, thresholds = roc_curve(true_labels, scores)\n",
        "\n",
        "# calculate the area under the ROC curve (AUC)\n",
        "auc_score = auc(fpr, tpr)\n",
        "\n",
        "# calculate d prime\n",
        "d_prime = np.abs(np.mean(scores[true_labels == 1]) - np.mean(scores[true_labels == 0])) / np.sqrt(0.5 * (np.std(scores[true_labels == 1]) ** 2 + np.std(scores[true_labels == 0]) ** 2))\n",
        "\n",
        "# Calculate the optimal operating point on the ROC curve\n",
        "optimal_idx = np.argmax(tpr - fpr)\n",
        "OPTROCPT = (fpr[optimal_idx], tpr[optimal_idx])\n",
        "\n",
        "# plot the ROC curve\n",
        "plt.plot(fpr, tpr, color='darkorange', lw=2, label='ROC curve (AUC = %0.4f, d\\' = %0.4f)' % (auc_score, d_prime))\n",
        "plt.plot(OPTROCPT[0], OPTROCPT[1], 'ro', label='Optimal Operating Point (fpr = %0.4f, tpr = %0.4f)' % (fpr[optimal_idx], tpr[optimal_idx]))\n",
        "plt.plot([0, 1], [0, 1], color='navy', lw=2, linestyle='--')\n",
        "plt.xlim([0.0, 1.0])\n",
        "plt.ylim([0.0, 1.0])\n",
        "plt.xlabel('False Positive Rate')\n",
        "plt.ylabel('True Positive Rate')\n",
        "plt.title('Receiver operating characteristic')\n",
        "plt.legend(loc=\"lower right\")\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 472
        },
        "id": "SvFF0t4Mh-g-",
        "outputId": "87b3f007-39a9-4915-c7aa-3f911767b463"
      },
      "execution_count": 29,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "juayZ-Afi9p4"
      },
      "execution_count": null,
      "outputs": []
    }
  ]
}