Coarse-to-fine Deep Video Coding with Hyperprior-guided Mode Prediction

TL;DR

This paper introduces a coarse-to-fine deep video coding framework that enhances motion compensation and employs hyperprior-guided mode prediction to improve compression efficiency, achieving state-of-the-art results without significant extra bit cost.

Contribution

It proposes a novel coarse-to-fine motion compensation framework and hyperprior-guided mode prediction methods that improve compression performance in deep video coding.

Findings

Achieves state-of-the-art performance on multiple datasets.

Improves motion compensation without increasing bit cost.

Effectively predicts block resolutions and residual skipping using hyperprior information.

Abstract

The previous deep video compression approaches only use the single scale motion compensation strategy and rarely adopt the mode prediction technique from the traditional standards like H.264/H.265 for both motion and residual compression. In this work, we first propose a coarse-to-fine (C2F) deep video compression framework for better motion compensation, in which we perform motion estimation, compression and compensation twice in a coarse to fine manner. Our C2F framework can achieve better motion compensation results without significantly increasing bit costs. Observing hyperprior information (i.e., the mean and variance values) from the hyperprior networks contains discriminant statistical information of different patches, we also propose two efficient hyperprior-guided mode prediction methods. Specifically, using hyperprior information as the input, we propose two mode prediction networks to respectively predict the optimal block resolutions for better motion coding and decide whether to skip residual information from each block for better residual coding without introducing additional bit cost while bringing negligible extra computation cost. Comprehensive experimental results demonstrate our proposed C2F video compression framework equipped with the new hyperprior-guided mode prediction methods achieves the state-of-the-art performance on HEVC, UVG and MCL-JCV datasets.