Histogram Layers for Neural Engineered Features:

Histogram Layers for Neural Engineered Features

Joshua Peeples, Salim Al Kharsa, Luke Saleh, and Alina Zare

Note: If this code is used, cite it: Joshua Peeples, Salim Al Kharsa, Luke Saleh, and Alina Zare. (2024, March 25). Advanced Vision and Learning Lab/Neural Engineered Features: Initial Release (Version v1.0). Zendo. https://doi.org/10.5281/zenodo.10810460

arXiv

BibTeX

In this repository, we provide the paper and code for the "Histogram Layers for Neural Engineered Features"

Installation Prerequisites

This code uses python, pytorch, and associated libraries listed in req.txt Please use Pytorch's website to download necessary packages. Download other dependencies from req.txt using pip install -r req.txt

Demo

Run demo.py in Python IDE (e.g., Spyder) or command line.

Main Functions

The runs using the following functions.

1. Intialize model

model = intialize_model(**Parameters)

2. Prepare dataset(s) for model

dataloaders_dict = Prepare_Dataloaders(**Parameters)

3. Train model

train_dict = train_model(**Parameters)

4. Test model

test_dict = test_model(**Parameters)

5. Histogram Methods

histogram_layer = NLBPLayer(**Parameters) histogram_layer = NEHDLayer(**Parameters)

Parameters

The parameters can be set in the following script:

Demo_Parameters.py

Inventory

https://github.com/Advanced-Vision-and-Learning-Lab/NEHD_NLBP

└── root dir
	├── demo.py   //Run this. Main demo file.
	├── Demo_Parameters.py // Parameters file for demo.
	├── Prepare_Data.py  // Load data for demo file.
	├── View_Results.py // Run this after demo to view saved results.
  	├── req.txt // Contains the requirements 
        ├── Datasets
        	├── PRMIDataset.py // Returns Index for PRMI dataset
        	├── Pytorch_Datasets.py // Return Index for Pytorch datasets
	└── Utils  //utility functions
		├── Base_Model.py // Returns the linear model passed through
		├── Compute_EHD.py // Returns the EHD algorithm layer
		├── Compute_FDR.py // Returns the FDR score
		├── Compute_LBP.py // Returns the LBP algorithm layer
		├── Compute_Sizes.py // Returns the output size
		├── Confusion_mats.py // Returns the confusion matrices visual
		├── Generate_Learning_Curves.py // Returns the learning curves both loss and accuracy
		├── Generate_Plots.py // Returns the plots of the Histograms and Kernels learned
		├── Generate_TSNE_visual.py // Returns the TSNE plots for test data
		├── Histogram_Model.py // Returns the Histogram Network
		├── NEHD.py // Returns the NEHD layer implementation
		├── NLBP.py // Returns the NLBP layer implementation
		├── Network_functions.py // Contains the functions called in main
		├── RBFHistogramPooling.py // Returns the Histogram Layer
		├── Save_Results.py // Saves the results into the directory
		├── pytorchtools.py // Contains early stopping functionality

License

This source code is licensed under the license found in the LICENSE file in the root directory of this source tree.

This product is Copyright (c) 2024 J. Peeples, and Salim Al Kharsa, and Luke Saleh, and Alina Zare. All rights reserved.

Citing Histogram Layers for Neural Engineered Features

If you use the code, please cite the following reference using the following entry.

Plain Text:

J.Peeples and S.Al Kharsa and L.Saleh and A.Zare, "Histogram Layers for Neural Engineered Features," in Review.

BibTex:

@article{peeples2024histogram,
  title={Histogram Layers for Neural Engineered Features},
  author={Peeples, Joshua and Kharsa, Salim Al and Saleh, Luke and Zare, Alina},
  journal={arXiv preprint arXiv:2403.17176},
  year={2024}
}