Brian9999/diff-reflection-separation

1---2base_model:3  - stabilityai/stable-diffusion-24pipeline_tag: image-to-image5library_name: diffusers6tags:7  - reflection-removal8  - reflection-separation9  - image-restoration10  - diffusion11  - stable-diffusion12  - cvpr202613---14 15# Reflection Separation from a Single Image via Joint Latent Diffusion16 17Pre-trained checkpoints for the CVPR 2026 paper. Given a single photo taken18through glass, the model jointly generates the **transmission** and19**reflection** layers.20 21- 🌐 Project page: https://brian90709.github.io/diff-reflection-separation/22- 💻 Code: https://github.com/Brian90709/diff-reflection-separation-code23- 📄 Paper: arXiv (coming soon)24 25[Zheng-Hui Huang](https://github.com/Brian90709), [Zhixiang Wang](https://lightchaserx.github.io), [Yu-Lun Liu](https://yulunalexliu.github.io), [Yung-Yu Chuang](https://www.csie.ntu.edu.tw/~cyy/)26 27## Files28 29| File | Size | Description |30| --- | --- | --- |31| `iter_016000/unet/diffusion_pytorch_model.bin` | ~3.5 GB | Trained layer-separation UNet. |32| `fuse_blocks.bin` | ~264 MB | CFW refiner for the VAE decoder. |33| `lrm/iter_008000/aux_net.bin` | ~1.3 MB | Latent composition module (LRM), used by `--optimization`. |34 35## Usage36 37Download the weights into `./checkpoints`, then follow the38[code repository](https://github.com/Brian90709/diff-reflection-separation-code):39 40```bash41huggingface-cli download Brian9999/diff-reflection-separation --repo-type model --local-dir ./checkpoints42```43 44## Citation45 46```bibtex47@inproceedings{huang2026reflection,48  title     = {Reflection Separation from a Single Image via Joint Latent Diffusion},49  author    = {Huang, Zheng-Hui and Wang, Zhixiang and Liu, Yu-Lun and Chuang, Yung-Yu},50  booktitle = {CVPR},51  year      = {2026}52}53```54