---
license: apache-2.0
---
# [CVPR2024 Oral] Image Processing GNN: Breaking Rigidity in Super-Resolution
Paper |
Project Page |
This is the official repo of our CVPR'24 paper **Image Processing GNN: Breaking Rigidity in Super-Resolution**. In the paper, we propose IPG: a Graph-based SR model that achieves outstanding performance on various SR benchmarks.
## News
7/6/2024: We opensourced the code & weights of IPG!
6/19/2024: Our work got the **Best Student Runner-up Award** of CVPR'24!🎉🎉
6/2/2024: We open-sourced [U-DiT](https://github.com/YuchuanTian/U-DiT), an efficient U-Net-style DiT variant.
## Weights & Visual Results
| Model | Scale | Urban100 | Weights | Visual Results |
| ----- | ----- | -------- | ------------------------------------------------------------ | ------------------------------------------------------------ |
| IPG | 2x | 34.48 | [🤗Link](https://huggingface.co/yuchuantian/IPG/blob/main/IPG_SRx2.pth) | [🤗Link](https://huggingface.co/yuchuantian/IPG/blob/main/IPG_srx2.zip) |
| IPG | 3x | 30.36 | [🤗Link](https://huggingface.co/yuchuantian/IPG/blob/main/IPG_SRx3.pth) | [🤗Link](https://huggingface.co/yuchuantian/IPG/blob/main/IPG_srx3.zip) |
| IPG | 4x | 28.13 | [🤗Link](https://huggingface.co/yuchuantian/IPG/blob/main/IPG_SRx4.pth) | [🤗Link](https://huggingface.co/yuchuantian/IPG/blob/main/IPG_srx4.zip) |