Convolutional Neural Networks for Non-iterative Reconstruction of Compressively Sensed Images

TL;DR

This paper introduces ReconNet, a deep neural network that performs real-time, non-iterative image reconstruction from compressive measurements, outperforming traditional methods especially at low measurement rates and in noisy conditions.

Contribution

The paper presents a novel end-to-end trained deep neural network for compressive sensing reconstruction, including variants with adversarial loss and joint measurement matrix learning.

Findings

ReconNet achieves higher PSNR than iterative algorithms at low measurement rates.

The network is robust to sensor noise in real-world data.

Reconstructed images retain semantic content useful for high-level tasks.

Abstract

Traditional algorithms for compressive sensing recovery are computationally expensive and are ineffective at low measurement rates. In this work, we propose a data driven non-iterative algorithm to overcome the shortcomings of earlier iterative algorithms. Our solution, ReconNet, is a deep neural network, whose parameters are learned end-to-end to map block-wise compressive measurements of the scene to the desired image blocks. Reconstruction of an image becomes a simple forward pass through the network and can be done in real-time. We show empirically that our algorithm yields reconstructions with higher PSNRs compared to iterative algorithms at low measurement rates and in presence of measurement noise. We also propose a variant of ReconNet which uses adversarial loss in order to further improve reconstruction quality. We discuss how adding a fully connected layer to the existing ReconNet architecture allows for jointly learning the measurement matrix and the reconstruction algorithm in a single network. Experiments on real data obtained from a block compressive imager show that our networks are robust to unseen sensor noise. Finally, through an experiment in object tracking, we show that even at very low measurement rates, reconstructions using our algorithm possess rich semantic content that can be used for high level inference.