Transform Network Architectures for Deep Learning based End-to-End Image/Video Coding in Subsampled Color Spaces

TL;DR

This paper explores deep learning architectures for end-to-end image/video coding in subsampled YUV 4:2:0 format, proposing a new transform network that outperforms existing methods and standard codecs like HEVC.

Contribution

It introduces a novel transform network architecture tailored for YUV 4:2:0 format, enhancing coding efficiency in deep learning based end-to-end image/video compression.

Findings

Proposed architecture achieves about 10% BD-rate reduction over HEVC intra coding.

Supports YUV 4:2:0 format, aligning with traditional video compression standards.

Outperforms naive extensions of RGB-based architectures in experiments.

Abstract

Most of the existing deep learning based end-to-end image/video coding (DLEC) architectures are designed for non-subsampled RGB color format. However, in order to achieve a superior coding performance, many state-of-the-art block-based compression standards such as High Efficiency Video Coding (HEVC/H.265) and Versatile Video Coding (VVC/H.266) are designed primarily for YUV 4:2:0 format, where U and V components are subsampled by considering the human visual system. This paper investigates various DLEC designs to support YUV 4:2:0 format by comparing their performance against the main profiles of HEVC and VVC standards under a common evaluation framework. Moreover, a new transform network architecture is proposed to improve the efficiency of coding YUV 4:2:0 data. The experimental results on YUV 4:2:0 datasets show that the proposed architecture significantly outperforms naive extensions of existing architectures designed for RGB format and achieves about 10% average BD-rate improvement over the intra-frame coding in HEVC.