Deep Mean-Shift Priors for Image Restoration

By Siavash Arjomand Bigdeli, Matthias Zwicker, Paolo Favaro, Meiguang Jin. 2017

147 citations recorded by the source corpus.

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.

Subfields: Biomedical Engineering, Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Biomedical Engineering

In Computer Vision and Pattern Recognition

Citations in the corpus

GAN Inversion: A Survey (2021, background): “There are also other deep priors developed for low-level restoration tasks like deep denoiser prior [201], [202] and TNRD [203].”
Blind image super-resolution with spatially variant degradations (2019, background): “Instead of considering a denoising prior, Bigdeli et al. [2017] propose to use a prior based on an estimate of natural image distribution.”
Deep Image Deblurring: A Survey (2022, background): “The learned priors from [7,46,154] are not always”
Learning to Optimize: A Primer and A Benchmark (2021, background): “[27; 26] modeled the statistical distribution of natural images as a Denoising Autoencoder (DAE) and constructed the optimization objective by maximizing the likelihood.”
Neural Nearest Neighbors Networks (2018, background): “Moreover, denoising has been recognized as a core module for density estimation [2] and serves as a sub-routine for more general image restoration tasks in a ﬂurry of recent work, e. g. [5, 36, 52].”
Exploiting Deep Generative Prior for Versatile Image Restoration and Manipulation (2020, background): “ors developed for low-level restoration tasks like deep denoiser prior [42,6] and TNRD [8], but competing with them is not our goal.”
Learning a Single Convolutional Super-Resolution Network for Multiple Degradations (2017, methodology): “An interesting line of CNN-based methods which can go beyond bicubic degradation adopt a CNN denoiser to solve SISR via model-based optimization framework [4, 34, 57].”
Deep Learning on Image Denoising: An overview (2019, unknown): “They included BSD400 [19] and Waterloo Exploration Database [133].”
Deep Plug-And-Play Super-Resolution for Arbitrary Blur Kernels (2019, methodology): “During the past few years, a flurry of plug-and-play works have been developed from the following aspects: 1) different variable splitting algorithms, such as halfquadratic splitting (HQS) algorithm [1], alternating direction method of multipliers (ADMM) algorithm [8], FISTA [4], and primal-dual alg”