Image Compressed Sensing Using Non-local Neural Network

TL;DR

This paper introduces NL-CSNet, a novel deep learning framework for image compressed sensing that leverages non-local self-similarity priors in both measurement and feature domains to improve reconstruction quality.

Contribution

The paper proposes a new neural network architecture incorporating non-local subnetworks and a specialized loss function for enhanced image reconstruction in compressed sensing.

Findings

NL-CSNet outperforms existing methods in reconstruction quality.

The framework maintains fast computational speed.

Utilizes non-local priors in measurement and feature domains for better results.

Abstract

Deep network-based image Compressed Sensing (CS) has attracted much attention in recent years. However, the existing deep network-based CS schemes either reconstruct the target image in a block-by-block manner that leads to serious block artifacts or train the deep network as a black box that brings about limited insights of image prior knowledge. In this paper, a novel image CS framework using non-local neural network (NL-CSNet) is proposed, which utilizes the non-local self-similarity priors with deep network to improve the reconstruction quality. In the proposed NL-CSNet, two non-local subnetworks are constructed for utilizing the non-local self-similarity priors in the measurement domain and the multi-scale feature domain respectively. Specifically, in the subnetwork of measurement domain, the long-distance dependencies between the measurements of different image blocks are established for better initial reconstruction. Analogically, in the subnetwork of multi-scale feature domain, the affinities between the dense feature representations are explored in the multi-scale space for deep reconstruction. Furthermore, a novel loss function is developed to enhance the coupling between the non-local representations, which also enables an end-to-end training of NL-CSNet. Extensive experiments manifest that NL-CSNet outperforms existing state-of-the-art CS methods, while maintaining fast computational speed.