@misc{indiciae50e45f2a54e1,
  title = {FoundIR-v2: Optimizing Pre-Training Data Mixtures for Image Restoration Foundation Model},
  author = {Xiang Chen and Jinshan Pan and Jiangxin Dong and Jian Yang and Jinhui Tang},
  year = {2025},
  url = {https://arxiv.org/abs/2512.09282},
  note = {Source identifier: 2512.09282}
}