AMPA-Net: Optimization-Inspired Attention Neural Network for Deep Compressed Sensing

TL;DR

AMPA-Net is a novel neural network that combines optimization algorithms with attention mechanisms to improve image compressed sensing reconstruction speed and accuracy, demonstrating superior performance on benchmark datasets.

Contribution

The paper introduces AMP-Net, a neural network inspired by the Approximate Message Passing algorithm, and enhances it with attention modules in AMPA-Net for better image reconstruction.

Findings

AMP-Net effectively fuses optimization algorithms with neural networks.

AMPA-Net with attention modules outperforms existing methods on benchmarks.

The proposed models achieve faster and more accurate image reconstruction.

Abstract

Compressed sensing (CS) is a challenging problem in image processing due to reconstructing an almost complete image from a limited measurement. To achieve fast and accurate CS reconstruction, we synthesize the advantages of two well-known methods (neural network and optimization algorithm) to propose a novel optimization inspired neural network which dubbed AMP-Net. AMP-Net realizes the fusion of the Approximate Message Passing (AMP) algorithm and neural network. All of its parameters are learned automatically. Furthermore, we propose an AMPA-Net which uses three attention networks to improve the representation ability of AMP-Net. Finally, We demonstrate the effectiveness of AMP-Net and AMPA-Net on four standard CS reconstruction benchmark data sets. Our code is available on https://github.com/puallee/AMPA-Net.