Abstract:Deep neural networks (DNNs) have achieved significant success in image restoration tasks by directly learning a powerful non-linear mapping from corrupted images to their latent clean ones. However, there still exist two major limitations for these deep learning (DL)-based methods. Firstly, the noises contained in real corrupted images are very complex, usually neglected and largely under-estimated in most current methods. Secondly, existing DL methods are mostly trained on one pre-assumed degradation process for all of the training image pairs, such as the widely used bicubic downsampling assumption in the image super-resolution task, inevitably leading to poor generalization performance when the true degradation does not match with such assumed one. To address these issues, we propose a unified generative model for the image restoration, which elaborately configures the degradation process from the latent clean image to the observed corrupted one. Specifically, different from most of current methods, the pixel-wisely non-i.i.d. Gaussian distribution, being with more flexibility, is adopted in our method to fit the complex real noises. Furthermore, the method is built on the general image degradation process, making it capable of adapting diverse degradations under one single model. Besides, we design a variational inference algorithm to learn all parameters involved in the proposed model with explicit form of objective loss. Specifically, beyond traditional variational methodology, two DNNs are employed to parameterize the posteriori distributions, one to infer the distribution of the latent clean image, and another to infer the distribution of the image noise. Extensive experiments demonstrate the superiority of the proposed method on three classical image restoration tasks, including image denoising, image super-resolution and JPEG image deblocking.

A Convolution Neural Network Based Variational Restoration Model

Variational Image Restoration Network.

Variational Image Restoration Network

An Overview of Image Restoration Based on Variational Regularization

Variational Deep Image Restoration

Guided Regularization and its Application for Image Restoration

A Generalized Hybrid Nonconvex Variational Regularization Model for Staircase Reduction in Image Restoration

A regularized restoration model based on geometrical features and noise evaluation

A Local Characteristic Image Restoration Based on Convolutional Neural Network

Revitalizing Convolutional Network for Image Restoration

Convolutional Neural Network Combined with Half-Quadratic Splitting Method for Image Restoration

Constrained Minimization Problem for Image Restoration Based on Non-Convex Hybrid Regularization

On Nonconvex and Nonsmooth Regularization for Contrast-Preserving Variational Signal and Image Restoration

Adaptive Super-Resolution Image Restoration Based on Regularization in Spatial Domain

Adaptive Nonconvex Nonsmooth Regularization for Image Restoration Based on Spatial Information

A Novel Image-Restoration Method Based on High-Order Total Variation Regularization Term

Impulse Noise Image Restoration Using Nonconvex Variational Model and Difference of Convex Functions Algorithm.

Learning local regularization for variational image restoration

A Novel Variable Exponent Non-Convex TVp,q(x) Model in Image Restoration

On a General Smoothly Truncated Regularization for Variational Piecewise Constant Image Restoration: Construction and Convergent Algorithms