--- license: bsd-3-clause library_name: pytorch tags: - image-denoising - self-supervised-learning - deepinv - arxiv:2603.25869 --- # Learning to Recorrupt (L2R) **Noise distribution agnostic self-supervised image denoising.** Brayan Monroy · Jorge Bacca · Julian Tachella [Paper](https://arxiv.org/abs/2603.25869) · [Code](https://github.com/bemc22/learning2recorrupt) · [Weights](https://huggingface.co/bemc22/learning2recorrupt/tree/main) L2R jointly learns a denoiser and a monotonic recorruptor using noisy images alone. ### Weights Eight RGB DRUNet checkpoints cover log-gamma, Laplace, correlated Gaussian, Poisson–Gaussian, and Poisson noise. [manifest.json](manifest.json) contains their settings and SHA-256 hashes. [mfcn_init.pth](mfcn_init.pth) initializes the monotonic network as a Gaussian identity map. ### Use From the code repository, with dependencies installed and test caches available: ```bash hf download bemc22/learning2recorrupt --local-dir pretrained python test.py --checkpoint-dir pretrained ``` ### Citation ```bibtex @article{monroy2026learning, title={Learning to Recorrupt: Noise Distribution Agnostic Self-Supervised Image Denoising}, author={Monroy, Brayan and Bacca, Jorge and Tachella, Juli{\'a}n}, journal={arXiv preprint arXiv:2603.25869}, year={2026} } ```