Hugging Face's logo

LuizScarlet
/
RAE-CoD

Image-to-Image
Model card Files
xet
 Community
1

RAE-CoD

This repository contains the model described in Rethinking Generative Image Compression at Extremely Low Bitrates.

Code: https://github.com/LuizScarlet/RAE-CoD

The model performs representation-space generative image compression, producing visually plausible reconstructions even at extremely low bitrates (down to 16 bits for a 256×256 image).

Downloads last month

-

Downloads are not tracked for this model. How to track
Inference Providers NEW
Image-to-Image
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Paper for LuizScarlet/RAE-CoD