DeepInverse: A Python package for solving imaging inverse problems with deep learning

2025/05/26 by Julián Tachella, Matthieu Terris, Tachella, Julián +50 · 16 citations
Mathematics · Computer Science · Engineering · #Statistical and numerical algorithms #Computational Physics and Python Applications #Sparse and Compressive Sensing Techniques

paper · pdf · doi:10.48550/arxiv.2505.20160

Abstract

DeepInverse is an open-source PyTorch-based library for solving imaging inverse problems. The library covers all crucial steps in image reconstruction from the efficient implementation of forward operators (e.g., optics, MRI, tomography), to the definition and resolution of variational problems and the design and training of advanced neural network architectures. In this paper, we describe the main functionality of the library and discuss the main design choices.

Cited by

• Self-Supervised Learning from Noisy and Incomplete Data
• CHEM: Estimating and Understanding Hallucinations in Deep Learning for Image Processing
• Explainable Learning Based Regularization of Inverse Problems
• Equivariant Deep Equilibrium Models for Imaging Inverse Problems
• Bayesian model selection and misspecification testing in imaging inverse problems only from noisy and partial measurements
• A Relaxed Gradient Step Denoiser for Splitting Methods in Poisson Inverse Problems
• Structured Random Models for Phase Retrieval with Optical Diffusers
• Learning Regularization Functionals for Inverse Problems: A Comparative Study
• LVTINO: LAtent Video consisTency INverse sOlver for High Definition Video Restoration
• Equivariant Splitting: Self-supervised learning from incomplete data
• UTOPY: Unrolling Algorithm Learning via Fidelity Homotopy for Inverse Problems
• A Computational Framework and Implementation of Implicit Priors in Bayesian Inverse Problems
• Stabilizing RED using the Koopman Operator
• Preconditioned Regularized Wasserstein Proximal Sampling
• How Diffusion Prior Landscapes Shape the Posterior in Blind Deconvolution
• MRpro - open PyTorch-based MR reconstruction and processing package

Related

• Deep Learning Techniques for Inverse Problems in Imaging
• Deep Decomposition Learning for Inverse Imaging Problems
• Deep Equilibrium Architectures for Inverse Problems in Imaging
• Understanding Untrained Deep Models for Inverse Problems: Algorithms and Theory
• Data-driven approaches to inverse problems