PyTorch implementation of model based inverse problems library.
🚧 Still Under Construction — Stay Tuned! 🚧
This repository will provide a modular framework for inverse problems, supporting
multiple optimization algorithms, data terms, and priors (both classical and deep).
Planned components:
- First-order solvers (PGD, ADMM, PDHG)
- Second-order solvers (Quasi-Newton, Newton-CG)
- Plug-and-Play & RED priors
- MRI, CT, and Ptychography forward models
- Image deblurring/inpainting/supersolution etc.
- Dynamic Contrast-Enhanced Imaging.
- Nonlinear Inverse Problems, e.g., optical diffraction tomography
- Computational Methods for 4(5)D Dynamic Imaging