On learning optimized reaction diffusion processes for effective image restoration

Abstract

For several decades, image restoration remains an active research topic in low-level computer vision and hence new approaches are constantly emerging. However, many recently proposed algorithms achieve state-of-the-art performance only at the expense of very high computation time, which clearly limits their practical relevance. In this work, we propose a simple but effective approach with both high computational efficiency and high restoration quality. We extend conventional nonlinear reaction diffusion models by several parametrized linear filters as well as several parametrized influence functions. We propose to train the parameters of the filters and the influence functions through a loss based approach. Experiments show that our trained nonlinear reaction diffusion models largely benefit from the training of the parameters and finally lead to the best reported performance on common test datasets for image restoration. Due to their structural simplicity, our trained models are highly efficient and are also well-suited for parallel computation on GPUs.

9 pages, 3 figures, 3 tables. CVPR2015 oral presentation together with the supplemental material of 13 pages, 8 pages (Notes on diffusion networks)

References in corpus (1)

• Insights into analysis operator learning: From patch-based sparse models to higher-order MRFs

Cited by in corpus (25)

• Beyond a Gaussian Denoiser: Residual Learning of Deep CNN for Image Denoising
• Deep Convolutional Neural Network for Inverse Problems in Imaging
• Deep Residual Learning for Compressed Sensing CT Reconstruction via Persistent Homology Analysis
• Real-world Noisy Image Denoising: A New Benchmark
• MR image reconstruction using deep density priors
• External Prior Guided Internal Prior Learning for Real-World Noisy Image Denoising
• Recurrent Inference Machines for Solving Inverse Problems
• Learning a Single Convolutional Super-Resolution Network for Multiple Degradations
• Multi-channel Weighted Nuclear Norm Minimization for Real Color Image Denoising
• Crafting a Toolchain for Image Restoration by Deep Reinforcement Learning
• On the Relation between Color Image Denoising and Classification
• Deep Gaussian Conditional Random Field Network: A Model-based Deep Network for Discriminative Denoising
• Meta-MgNet: Meta Multigrid Networks for Solving Parameterized Partial Differential Equations
• One-to-Many Network for Visually Pleasing Compression Artifacts Reduction
• Toward Designing Convergent Deep Operator Splitting Methods for Task-specific Nonconvex Optimization
• Generalized Deep Image to Image Regression
• Fast and Accurate Poisson Denoising with Optimized Nonlinear Diffusion
• Deep Learning with Inaccurate Training Data for Image Restoration
• Beyond Deep Residual Learning for Image Restoration: Persistent Homology-Guided Manifold Simplification
• KRNET: Image Denoising with Kernel Regulation Network
• Accelerating GMM-based patch priors for image restoration: Three ingredients for a 100$\times$ speed-up
• Generic 3D Convolutional Fusion for image restoration
• Are good local minima wide in sparse recovery?
• cvpaper.challenge in 2016: Futuristic Computer Vision through 1,600 Papers Survey
• Learning Collaborative Generation Correction Modules for Blind Image Deblurring and Beyond