CS-MCNet:A Video Compressive Sensing Reconstruction Network with Interpretable Motion Compensation

TL;DR

CS-MCNet is a deep neural network for high-quality, real-time video compressive sensing reconstruction that uses interpretable motion compensation and algorithm unrolling, outperforming traditional methods in speed and accuracy.

Contribution

Introduces CS-MCNet with explicit multi-hypothesis motion compensation and interpretable architecture for improved video reconstruction performance.

Findings

Achieves 22dB PSNR at 64x compression ratio, outperforming state-of-the-art by 4-9%.

Reconstruction is up to 1000 times faster than traditional iterative methods.

Network enables real-time processing of compressed videos.

Abstract

In this paper, a deep neural network with interpretable motion compensation called CS-MCNet is proposed to realize high-quality and real-time decoding of video compressive sensing. Firstly, explicit multi-hypothesis motion compensation is applied in our network to extract correlation information of adjacent frames(as shown in Fig. 1), which improves the recover performance. And then, a residual module further narrows down the gap between reconstruction result and original signal. The overall architecture is interpretable by using algorithm unrolling, which brings the benefits of being able to transfer prior knowledge about the conventional algorithms. As a result, a PSNR of 22dB can be achieved at 64x compression ratio, which is about 4% to 9% better than state-of-the-art methods. In addition, due to the feed-forward architecture, the reconstruction can be processed by our network in real time and up to three orders of magnitude faster than traditional iterative methods.