DeepInverse: a PyTorch library for imaging with deep learning#

DeepInverse is a PyTorch-based library for solving imaging inverse problems with deep learning.

Featuring

Large collection of predefined imaging operators (MRI, CT, deblurring, inpainting, etc.)
Training losses for inverse problems (self-supervised learning, regularization, etc.).
Many pretrained deep denoisers which can be used for plug-and-play restoration.
Framework for building datasets for inverse problems.
Easy-to-build unfolded architectures (ADMM, forward-backward, deep equilibrium, etc.).
Diffusion algorithms for image restoration and uncertainty quantification (Langevin, diffusion, etc.).
A large number of well-explained examples, from basics to state-of-the-art methods.

Citation

If you use DeepInverse in your research, please cite the following paper (available on arXiv):

@software{tachella2025deepinverse,
      title={DeepInverse: A Python package for solving imaging inverse problems with deep learning},
      author={Julián Tachella and Matthieu Terris and Samuel Hurault and Andrew Wang and Dongdong Chen and Minh-Hai Nguyen and Maxime Song and Thomas Davies and Leo Davy and Jonathan Dong and Paul Escande and Johannes Hertrich and Zhiyuan Hu and Tobías I. Liaudat and Nils Laurent and Brett Levac and Mathurin Massias and Thomas Moreau and Thibaut Modrzyk and Brayan Monroy and Sebastian Neumayer and Jérémy Scanvic and Florian Sarron and Victor Sechaud and Georg Schramm and Romain Vo and Pierre Weiss},
      year={2025},
      eprint={2505.20160},
      archivePrefix={arXiv},
      primaryClass={eess.IV},
      url={https://arxiv.org/abs/2505.20160},
}

Star history

Lead Developers

Julian Tachella, Dongdong Chen, Samuel Hurault, Matthieu Terris and Andrew Wang.