The MAP model is a famous framework in the field of image restoration, and recent years have also witnessed the unprecedented success of CNNs in image denoising and super-resolution. In this talk, we intend to introduce some works on CNNs by referring to MAP inference in low level vision tasks: (1) design of a CNN denoiser, (2) extension of CNN denoisers for image restoration, and (3) other insights on CNN-based models delivered by MAP inference.

This presentation is part of Minisymposium “MS52 - A Denoiser Can Do Much More Than Just... Denoising (2 parts)”
organized by: Yaniv Romano (Technion - Israel Institute of Technology) , Peyman Milanfar (Google Research) , Michael Elad (The Technion - Israel Institute of Technology) .

Authors:

Wangmeng Zuo (Harbin Institute of Technology)

Lei Zhang (Hong Kong Polytechnic University)

Kai Zhang (Harbin Institute of Technology)

Shuhang Gu (The Hong Kong Polytechnic University)

Keywords:

bayesian methods, image reconstruction, inverse problems