Learning for Video Compression

TL;DR

This paper introduces PixelMotionCNN, a novel neural network architecture that models spatiotemporal coherence for learning-based video compression, achieving performance comparable to H.264 without complex entropy coding.

Contribution

The paper proposes PixelMotionCNN, a new neural network framework that incorporates motion extension and hybrid prediction for improved learning-based video compression.

Findings

Outperforms MPEG-2 in compression efficiency

Achieves comparable results with H.264

Demonstrates the potential of learning-based schemes for future video coding

Abstract

One key challenge to learning-based video compression is that motion predictive coding, a very effective tool for video compression, can hardly be trained into a neural network. In this paper we propose the concept of PixelMotionCNN (PMCNN) which includes motion extension and hybrid prediction networks. PMCNN can model spatiotemporal coherence to effectively perform predictive coding inside the learning network. On the basis of PMCNN, we further explore a learning-based framework for video compression with additional components of iterative analysis/synthesis, binarization, etc. Experimental results demonstrate the effectiveness of the proposed scheme. Although entropy coding and complex configurations are not employed in this paper, we still demonstrate superior performance compared with MPEG-2 and achieve comparable results with H.264 codec. The proposed learning-based scheme provides a possible new direction to further improve compression efficiency and functionalities of future video coding.