Neural Proximal Gradient Descent for Compressive Imaging

Abstract

Recovering high-resolution images from limited sensory data typically leads to a serious ill-posed inverse problem, demanding inversion algorithms that effectively capture the prior information. Learning a good inverse mapping from training data faces severe challenges, including: (i) scarcity of training data; (ii) need for plausible reconstructions that are physically feasible; (iii) need for fast reconstruction, especially in real-time applications. We develop a successful system solving all these challenges, using as basic architecture the recurrent application of proximal gradient algorithm. We learn a proximal map that works well with real images based on residual networks. Contraction of the resulting map is analyzed, and incoherence conditions are investigated that drive the convergence of the iterates. Extensive experiments are carried out under different settings: (a) reconstructing abdominal MRI of pediatric patients from highly undersampled Fourier-space data and (b) superresolving natural face images. Our key findings include: 1. a recurrent ResNet with a single residual block unrolled from an iterative algorithm yields an effective proximal which accurately reveals MR image details. 2. Our architecture significantly outperforms conventional non-recurrent deep ResNets by 2dB SNR; it is also trained much more rapidly. 3. It outperforms state-of-the-art compressed-sensing Wavelet-based methods by 4dB SNR, with 100x speedups in reconstruction time.

arXiv admin note: text overlap with arXiv:1711.10046

Cited by in corpus (26)

• On instabilities of deep learning in image reconstruction - Does AI come at a cost?
• Self-Supervised Learning of Physics-Guided Reconstruction Neural Networks without Fully-Sampled Reference Data
• Understanding and mitigating gradient pathologies in physics-informed neural networks
• Learning to Optimize: A Primer and A Benchmark
• Wasserstein GANs for MR Imaging: from Paired to Unpaired Training
• Deep Residual Autoencoders for Expectation Maximization-inspired Dictionary Learning
• Learning Sampling and Model-Based Signal Recovery for Compressed Sensing MRI
• Deep Unfolding with Normalizing Flow Priors for Inverse Problems
• Neumann Networks for Inverse Problems in Imaging
• Optimization methods for MR image reconstruction (long version)
• Theoretical Interpretation of Learned Step Size in Deep-Unfolded Gradient Descent
• Compressed Sensing: From Research to Clinical Practice with Data-Driven Learning
• Compressed MRI Reconstruction Exploiting a Rotation-Invariant Total Variation Discretization
• Convex Regularization Behind Neural Reconstruction
• TFPnP: Tuning-free Plug-and-Play Proximal Algorithm with Applications to Inverse Imaging Problems
• Feasibility-based Fixed Point Networks
• A Design Space Study for LISTA and Beyond
• Compressive MR Fingerprinting reconstruction with Neural Proximal Gradient iterations
• Equivariant Imaging: Learning Beyond the Range Space
• Deep Decomposition Learning for Inverse Imaging Problems
• Improving Generalizability in Limited-Angle CT Reconstruction with Sinogram Extrapolation
• Learning A 3D-CNN and Transformer Prior for Hyperspectral Image Super-Resolution
• Compressed Sensing via Measurement-Conditional Generative Models
• Non-Convex Compressed Sensing with Training Data
• Convolutional dictionary learning based auto-encoders for natural exponential-family distributions
• Reconstructing unseen modalities and pathology with an efficient Recurrent Inference Machine