Compressive Imaging using Approximate Message Passing and a Markov-Tree Prior

Abstract

We propose a novel algorithm for compressive imaging that exploits both the sparsity and persistence across scales found in the 2D wavelet transform coefficients of natural images. Like other recent works, we model wavelet structure using a hidden Markov tree (HMT) but, unlike other works, ours is based on loopy belief propagation (LBP). For LBP, we adopt a recently proposed "turbo" message passing schedule that alternates between exploitation of HMT structure and exploitation of compressive-measurement structure. For the latter, we leverage Donoho, Maleki, and Montanari's recently proposed approximate message passing (AMP) algorithm. Experiments with a large image database suggest that, relative to existing schemes, our turbo LBP approach yields state-of-the-art reconstruction performance with substantial reduction in complexity.

References in corpus (1)

• Compressive Imaging using Approximate Message Passing and a Markov-Tree Prior

Cited by in corpus (44)

• Statistical physics of inference: Thresholds and algorithms
• Compressive Imaging using Approximate Message Passing and a Markov-Tree Prior
• Dynamic Compressive Sensing of Time-Varying Signals via Approximate Message Passing
• Capacity-achieving Sparse Superposition Codes via Approximate Message Passing Decoding
• Measure What Should be Measured: Progress and Challenges in Compressive Sensing
• Approximate message-passing decoder and capacity-achieving sparse superposition codes
• Compressive Imaging via Approximate Message Passing with Image Denoising
• On Convergence of Approximate Message Passing
• On asymptotic structure in compressed sensing
• A Unified Bayesian Inference Framework for Generalized Linear Models
• Joint Pilot Optimization, Target Detection and Channel Estimation for Integrated Sensing and Communication Systems
• Near optimal compressed sensing without priors: Parametric SURE Approximate Message Passing
• Hybrid Approximate Message Passing
• Breaking the coherence barrier: A new theory for compressed sensing
• ReconNet: Non-Iterative Reconstruction of Images from Compressively Sensed Random Measurements
• Low Complexity Sparse Bayesian Learning Using Combined BP and MF with a Stretched Factor Graph
• Forest Sparsity for Multi-channel Compressive Sensing
• Optimal Data Detection in Large MIMO
• Image Restoration and Reconstruction using Variable Splitting and Class-adapted Image Priors
• Concise Derivation of Complex Bayesian Approximate Message Passing via Expectation Propagation
• Asymptotic Analysis of Complex LASSO via Complex Approximate Message Passing (CAMP)
• Information-Theoretically Optimal Compressed Sensing via Spatial Coupling and Approximate Message Passing
• Signal Recovery in Unions of Subspaces with Applications to Compressive Imaging
• Accurate Prediction of Phase Transitions in Compressed Sensing via a Connection to Minimax Denoising
• Exact Reconstruction Conditions for Regularized Modified Basis Pursuit
• Fast Iteratively Reweighted Least Squares Algorithms for Analysis-Based Sparsity Reconstruction
• Approximate Message Passing with Nearest Neighbor Sparsity Pattern Learning
• Tree-Structure Bayesian Compressive Sensing for Video
• Bayesian Optimal Approximate Message Passing to Recover Structured Sparse Signals
• Compressed sensing with local structure: uniform recovery guarantees for the sparsity in levels class
• Expectation-Maximization-Aided Hybrid Generalized Expectation Consistent for Sparse Signal Reconstruction
• Fixed Points of Generalized Approximate Message Passing with Arbitrary Matrices
• Inference for Generalized Linear Models via Alternating Directions and Bethe Free Energy Minimization
• Capacity Optimality of AMP in Coded Systems
• Sample Distortion for Compressed Imaging
• A Max-Product EM Algorithm for Reconstructing Markov-tree Sparse Signals from Compressive Samples
• Regularization by Denoising: Clarifications and New Interpretations
• An Overview of Multi-Processor Approximate Message Passing
• Computationally Efficient Sparse Bayesian Learning via Generalized Approximate Message Passing
• Bayesian Convolutional Neural Networks for Compressed Sensing Restoration
• Optimal Number of Measurements in a Linear System with Quadratically Decreasing SNR
• From compression to compressed sensing
• Multiscale Shrinkage and Lévy Processes
• Radar Imaging by Sparse Optimization Incorporating MRF Clustering Prior