Cascaded Reconstruction Network for Compressive image sensing

TL;DR

This paper introduces two cascaded deep learning-based reconstruction networks for compressive image sensing, significantly improving reconstruction quality and efficiency over existing algorithms, with one network outperforming the other by over 1 dB.

Contribution

It proposes two novel cascaded neural network architectures tailored for different CS sampling methods, enhancing reconstruction accuracy and computational efficiency.

Findings

Both networks outperform state-of-the-art CS reconstruction algorithms.

ASRNet achieves over 1 dB higher PSNR than CSRNet.

The methods reduce reconstruction time compared to traditional algorithms.

Abstract

The theory of compressed sensing (CS) has been successfully applied to image compression in the past few years, whose traditional iterative reconstruction algorithm is time-consuming. However, it has been reported deep learning-based CS reconstruction algorithms could greatly reduce the computational complexity. In this paper, we propose two efficient structures of cascaded reconstruction networks corresponding to two different sampling methods in CS process. The first reconstruction network is a compatibly sampling reconstruction network (CSRNet), which recovers an image from its compressively sensed measurement sampled by a traditional random matrix. In CSRNet, deep reconstruction network module obtains an initial image with acceptable quality, which can be further improved by residual network module based on convolutional neural network. The second reconstruction network is adaptively sampling reconstruction network (ASRNet), by matching automatically sampling module with corresponding residual reconstruction module. The experimental results have shown that the proposed two reconstruction networks outperform several state-of-the-art compressive sensing reconstruction algorithms. Meanwhile, the proposed ASRNet can achieve more than 1 dB gain, as compared with the CSRNet.