-
-
Notifications
You must be signed in to change notification settings - Fork 13
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
feat: add AI conversion with a new method + minore fixies #20
Conversation
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Some minor fixes and I'm confident in your skill in creating the model.py
Co-authored-by: GabrielePuliti <12409541+Wabri@users.noreply.github.com>
Co-authored-by: GabrielePuliti <12409541+Wabri@users.noreply.github.com>
Co-authored-by: GabrielePuliti <12409541+Wabri@users.noreply.github.com>
Co-authored-by: GabrielePuliti <12409541+Wabri@users.noreply.github.com>
Addressed all the comment. At the moment I'd like to avoid throw an exception if we receive more than 6 colors in input as we need to collect some data from our demo website to improve the model |
@Wabri please re-review :) |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
LGTM 🥇 Lets go to the moon!
Let's gooooo ! |
This PR will introduce a new conversion method using the PaletteNet implementation in a six input color palette.
It's our first try to jump in a conversion - recolor palette algorithm.
It's been a long journey since the first iteration here. Back in 2020 we thought of implementing a GAN but it was too hard for us to implementing it.
Nowadays the tools are super powerful and easy to use (pytoch 2.2 it's crazy).
Thans to the researchers who made the PaletteNet, to the guys who implemented it: it enable us to make this PR.
It's not perfect and I think the linear algorithm it's better in some results BUT the AI model is working great for the wallpaper: it's giving back a really high definition image (no data loss, pixel quality).
Within this PR come also some minor fixies such as video in readme fix and others.
We're going to have a major release on this.
Closes #21
Closes #19
Closes #15
Closes #14
As a demo we're going to add the new method in our demo website and API on @Schrodinger-Hat/imagegonord