Deep Mean-Shift Priors for Image Restoration

Abstract

In this paper we introduce a natural image prior that directly represents a Gaussian-smoothed version of the natural image distribution. We include our prior in a formulation of image restoration as a Bayes estimator that also allows us to solve noise-blind image restoration problems. We show that the gradient of our prior corresponds to the mean-shift vector on the natural image distribution. In addition, we learn the mean-shift vector field using denoising autoencoders, and use it in a gradient descent approach to perform Bayes risk minimization. We demonstrate competitive results for noise-blind deblurring, super-resolution, and demosaicing.

NIPS 2017

References in corpus (4)

• Beyond a Gaussian Denoiser: Residual Learning of Deep CNN for Image Denoising
• Trainable Nonlinear Reaction Diffusion: A Flexible Framework for Fast and Effective Image Restoration
• Learning Deep CNN Denoiser Prior for Image Restoration
• The Little Engine that Could: Regularization by Denoising (RED)

Cited by in corpus (12)

• Test-Time Adaptable Neural Networks for Robust Medical Image Segmentation
• Learning to Optimize: A Primer and A Benchmark
• Fixed-Point and Objective Convergence of Plug-and-Play Algorithms
• Neural Nearest Neighbors Networks
• Deep Learning on Image Denoising: An overview
• Disentangling Factors of Variation by Mixing Them
• Non-Local ConvLSTM for Video Compression Artifact Reduction
• GridFace: Face Rectification via Learning Local Homography Transformations
• Scalable Plug-and-Play ADMM with Convergence Guarantees
• MeanShift++: Extremely Fast Mode-Seeking With Applications to Segmentation and Object Tracking
• Infusing Learned Priors into Model-Based Multispectral Imaging
• Boosting the Performance of Plug-and-Play Priors via Denoiser Scaling