FISTA-Net

A model-based deep learning network for inverse problem in imaging

General Framework

Steps:

• Run 'CalculateW.m' to get the learned weights (refer to Liu et al. 2019);
• _Competitive methods: 'M1LapReg.py', 'M2TVFISTA.m', 'M3FBPConv.py', 'M4ISTANet.py';
• Proposed: 'M5FISTANet.py' (without learned matrix); 'M5FISTANetPlus.py' (with learned matrix);

Note These are example codes for the EMT experiment. As for sparse-view CT, additional libraries like the radon transform operator should be included.