Top AI Repos — open-source AI, indexed and scored
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
A diffusion model to colorize black and white images
| Date | Stars |
|---|---|
| 2026-07-31 | 773 |
| 2026-08-02 | 773 |
| 2026-08-06 | 773 |
Today
— stars today
This week
— stars this week
This month
— stars this month
Momentum
0.0
growth rate 0.00%/day
# Color Diffusion Using diffusion models to colorize black and white images. <div> <img src="https://github.com/ErwannMillon/Color-diffusion/blob/main/visualization/inference/total_1.gif" width="128" height="128"/> <img src="https://github.com/ErwannMillon/Color-diffusion/blob/main/visualization/inference/total_2.gif" width="128" height="128"/> <img src="https://github.com/ErwannMillon/Color-diffusion/blob/main/visualization/inference/total_3.gif" width="128" height="128"/> <img src="https://github.com/ErwannMillon/Color-diffusion/blob/main/visualization/inference/total_4.gif" width="128" height="128"/> <img src="https://github.com/ErwannMillon/Color-diffusion/blob/main/visualization/inference/total_8.gif" width="128" height="128"/> <img src="https://github.com/ErwannMillon/Color-diffusion/blob/main/visualization/inference/total_90.gif" width="128" height="128"/> </div> ## Overview This project is a simple example of how we can use diffusion models to colorize black and white images. This implementation uses the LAB color space, a 3 channel alternative to the RGB color space. The "L" (Lightness) channel in this space is equivalent to a greyscale image: it represents the luminous intensity of each pixel. The two other channels are used to represent the color of each pixel. To train the model, we first load color images and convert them to LAB. Then, we add noise only to the color channels, keeping the L channel constant. The model gets this channel "for free" because it doesn't need to learn how to predict the greyscale image: it is always known at train and test time. <br></br> <figure> <div> <img src="https://github.com/ErwannMillon/Color-diffusion/blob/main/visualization/forward_diff.gif" width="128" height="128" /> </div> <figcaption align = "center"><b>Forward Diffusion Process</b></figcaption> </figure> <br></br> *Note that we actually don't need to go through all of the steps of the diffusion process to get to timestep t. Our forward diffusion process is non-Markovian, but the entire diffusion process is shown for illustrative purposes* The model is a UNet that takes a 3 channel LAB input (the ground-truth greyscale channel concatenated with noised AB channels) and outputs a 2 channel prediction of the color noise. <br></br> <figure> <div> <img src="https://github.com/ErwannMillon/Color-diffusion/blob/main/visualization/train/total1.gif" width="128" height="128"/> <img src="https://github.com/ErwannMillon/Color-diffusion/blob/main/visualization/train/total2.gif" width="128" height="128"/> <img src="https://github.com/ErwannMillon/Color-diffusion/blob/main/visualization/train/total3.gif" width="128" height="128"/> <img src="https://github.com/ErwannMillon/Color-diffusion/blob/main/visualization/train/total4.gif" width="128" height="128"/> </div> <figcaption align = "center"><b>Forward diffusion and denoising at train time</b></figcaption> </figure> <br></br> In addition to receiving the greyscale channel as input, the UNet is also conditioned on features extracted from the greyscale channel. Intermediate feature maps from an encoder (implemented as the first half of a UNet) are concatenated with the features of the main denoising UNet throughout the downsampling stage of the forward pass. ## Future Work / Ideas This was just a quick proof of concept to satisfy my curiosity and get a feel for training diffusion models from scratch, so the results are very basic. There are many ways this project could be improved, such as: - Using pretrained face recognition networks like ArcFace or FaceNet as feature extractors to get the conditioning features - Implementing cross attention on the embeddings - Pretraining the greyscale feature extractor as the encoder stage of a greyscale autoencoder ## References A lot of code for the dataset and LAB color operations was adapted from moein-sharitania's colorization project, which used Conditional GANs https://github.com/moein-shariatnia/Deep-Learning I implemented optional
Excerpt of 4,684 characters
Read on GitHub
Erwann Millon · Krea AI · United States
16
Would you bet a product on this? Bounded 0–100 and slow moving.
matched fp:ced4bde1facaa916, desc:diffusion model