Deep regularization networks for inverse problems with noisy operators

Fatemeh Pourahmadian, Yang Xu

A supervised learning approach is proposed for regularization of large inverse problems where the main operator is built from noisy data. This is germane to superresolution imaging…