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.

Riemannian Loss for Image Restoration

Loss Functions for Image Restoration with Neural Networks

Near-lossless L-infinity Constrained Multi-rate Image Decompression Via Deep Neural Network

Loss Functions for Neural Networks for Image Processing.

A CT Image Feature Space (CTIS) Loss for Restoration with Deep Learning-Based Methods

Near-Lossless ℓ∞-Constrained Image Decompression Via Deep Neural Network.

A Novel Loss Function for Optical and SAR Image Matching: Balanced Positive and Negative Samples

Near-lossless ` ∞-constrained Multi-rate Image Decompression via Deep Neural Network

Projected Distribution Loss for Image Enhancement

Variational Image Restoration Network

Variational Image Restoration Network.

Deep unfolding of a proximal interior point method for image restoration

LuminanceL1Loss: A loss function which measures percieved brightness and colour differences

LIR: A Lightweight Baseline for Image Restoration

Random Weights Networks Work as Loss Prior Constraint for Image Restoration

Variational Deep Image Restoration

Robust image restoration via adaptive low-rank approximation and joint kernel regression.

Group sparsity residual constraint model with weighted log-sum penalty for image restoration

Attention-Based Real Image Restoration

Robust Image Restoration via Reweighted Low-Rank Matrix Recovery

From Rank Estimation to Rank Approximation: Rank Residual Constraint for Image Restoration