Content-Aware Scalable Deep Compressed Sensing

To more efficiently address image compressed sensing (CS) problems, we present a novel content-aware scalable network dubbed CASNet which collectively achieves adaptive sampling rate allocation, fine granular scalability and high-quality reconstruction. We first adopt a data-driven saliency detector to evaluate the importance of different image regions and propose a saliency-based block ratio aggregation (BRA) strategy for sampling rate allocation. A unified learnable generating matrix is then developed to produce sampling matrix of any CS ratio with an ordered structure. Being equipped with the optimization-inspired recovery subnet guided by saliency information and a multi-block training scheme preventing blocking artifacts, CASNet jointly reconstructs the image blocks sampled at various sampling rates with one single model. To accelerate training convergence and improve network robustness, we propose an SVD-based initialization scheme and a random transformation enhancement (RTE) strategy, which are extensible without introducing extra parameters. All the CASNet components can be combined and learned end-to-end. We further provide a four-stage implementation for evaluation and practical deployments. Experiments demonstrate that CASNet outperforms other CS networks by a large margin, validating the collaboration and mutual supports among its components and strategies. Codes are available at https://github.com/Guaishou74851/CASNet.

I. Introduction

Compressed sensing (CS) is a novel paradigm that requires much fewer measurements than the Nyquist sampling for signal acquisition and restoration [1], [2]. For the signal , it conducts the sampling process to obtain the measurements , where with is a given sampling matrix, and the CS ratio (or sampling rate) is defined as . Since it is hardware-friendly and has great potentials of improving sampling speed with high recovery accuracy, many applications have been developed including single-pixel imaging [3], [4], magnetic resonance imaging (MRI) [5], [6], sparse-view CT [7], etc. In this work, we focus on the typical block-based (or block-diagonal) image CS problem [8]–[10] that divides the high-dimensional natural image into non-overlapped blocks and obtains measurements block-by-block with a small fixed sampling matrix for the subsequent reconstruction.