Learned Convolutional Sparse Coding

Abstract

We propose a convolutional recurrent sparse auto-encoder model. The model consists of a sparse encoder, which is a convolutional extension of the learned ISTA (LISTA) method, and a linear convolutional decoder. Our strategy offers a simple method for learning a task-driven sparse convolutional dictionary (CD), and producing an approximate convolutional sparse code (CSC) over the learned dictionary. We trained the model to minimize reconstruction loss via gradient decent with back-propagation and have achieved competitive results to KSVD image denoising and to leading CSC methods in image inpainting requiring only a small fraction of their run-time.

References in corpus (2)

• Convolutional Dictionary Learning: A Comparative Review and New Algorithms
• Optimization Methods for Convolutional Sparse Coding

Cited by in corpus (29)

• Learning to Optimize: A Primer and A Benchmark
• Deep Learning Methods for Solving Linear Inverse Problems: Research Directions and Paradigms
• Learned SPARCOM: Unfolded Deep Super-Resolution Microscopy
• Deep Residual Autoencoders for Expectation Maximization-inspired Dictionary Learning
• CDLNet: Noise-Adaptive Convolutional Dictionary Learning Network for Blind Denoising and Demosaicing
• Hybrid ISTA: Unfolding ISTA With Convergence Guarantees Using Free-Form Deep Neural Networks
• Structured LISTA for Multidimensional Harmonic Retrieval
• Frank-Wolfe Network: An Interpretable Deep Structure for Non-Sparse Coding
• Deep Convolutional Neural Network for Multi-modal Image Restoration and Fusion
• Mixed X-Ray Image Separation for Artworks with Concealed Designs
• GroupCDL: Interpretable Denoising and Compressed Sensing MRI via Learned Group-Sparsity and Circulant Attention
• SSFN -- Self Size-estimating Feed-forward Network with Low Complexity, Limited Need for Human Intervention, and Consistent Behaviour across Trials
• Deep Convolutional Sparse Coding Networks for Image Fusion
• Efficient and Model-Based Infrared and Visible Image Fusion Via Algorithm Unrolling
• DuRIN: A Deep-unfolded Sparse Seismic Reflectivity Inversion Network
• Image Super-Resolution via RL-CSC: When Residual Learning Meets Convolutional Sparse Coding
• SINET: Sparsity-driven Interpretable Neural Network for Underwater Image Enhancement
• CDLNet: Robust and Interpretable Denoising Through Deep Convolutional Dictionary Learning
• Feasibility-based Fixed Point Networks
• Multimodal Deep Unfolding for Guided Image Super-Resolution
• On The Relationship Between Universal Adversarial Attacks And Sparse Representations
• CRNet: Image Super-Resolution Using A Convolutional Sparse Coding Inspired Network
• NuSPAN: A Proximal Average Network for Nonuniform Sparse Model -- Application to Seismic Reflectivity Inversion
• Deeply-Sparse Signal rePresentations ($\text{D}{\text{S}}^2\text{P}$)
• Learned Interpretable Residual Extragradient ISTA for Sparse Coding
• Interpretable Deep Multimodal Image Super-Resolution
• Convolutional dictionary learning based auto-encoders for natural exponential-family distributions
• SpotNet - Learned iterations for cell detection in image-based immunoassays
• Scalable Convolutional Dictionary Learning with Constrained Recurrent Sparse Auto-encoders