Motion Free B-frame Coding for Neural Video Compression

TL;DR

This paper introduces a kernel-based motion-free neural video compression method that enhances efficiency and visual quality by removing motion estimation and compensation, outperforming state-of-the-art models on multiple datasets.

Contribution

It presents a novel motion-free approach that reduces computational complexity and improves visual quality in neural video compression, addressing limitations of traditional architectures.

Findings

Outperforms SOTA neural video compression networks on HEVC-class B dataset.

Achieves comparable quality with significantly smaller model size.

Reduces computational complexity by eliminating motion estimation and compensation.

Abstract

Typical deep neural video compression networks usually follow the hybrid approach of classical video coding that contains two separate modules: motion coding and residual coding. In addition, a symmetric auto-encoder is often used as a normal architecture for both motion and residual coding. In this paper, we propose a novel approach that handles the drawbacks of the two typical above-mentioned architectures, we call it kernel-based motion-free video coding. The advantages of the motion-free approach are twofold: it improves the coding efficiency of the network and significantly reduces computational complexity thanks to eliminating motion estimation, motion compensation, and motion coding which are the most time-consuming engines. In addition, the kernel-based auto-encoder alleviates blur artifacts that usually occur with the conventional symmetric autoencoder. Consequently, it improves the visual quality of the reconstructed frames. Experimental results show the proposed framework outperforms the SOTA deep neural video compression networks on the HEVC-class B dataset and is competitive on the UVG and MCL-JCV datasets. In addition, it generates high-quality reconstructed frames in comparison with conventional motion coding-based symmetric auto-encoder meanwhile its model size is much smaller than that of the motion-based networks around three to four times.