Net is defined over a data flow graph, which is derived from the iterative pro-
cedures in Alternating Direction Method of Multipliers (ADMM) algorithm for
optimizing a CS-based MRI model. In the training phase, all parameters of the
net, e.g., image transforms, shrinkage functions, etc., are discriminatively trained
end-to-end using L-BFGS algorithm. In the testing phase, it has computational
overhead similar to ADMM but uses optimized parameters learned from the train-
ing data for CS-based reconstruction task. Experiments on MRI image reconstruc-
tion under different sampling ratios in k-space demonstrate that it significantly
improves the baseline ADMM algorith