The purpose of this paper is to provide an overview of VVC, with HEVC as a reference design, and to report performance evaluations of VVC compared to HEVC. 

Abstract

The Versatile Video Coding (VVC) is the most recent video coding standard jointly developed by MPEG (ISO/IEC) and VCEG (ITU-T) in the JVET (Joint Video Experts Team). The VVC Final Draft International Standard was issued in mid-2020. VVC can be considered as the state-of-the-art video coding standard, with an estimated bitrate gain around 40% versus the High Efficiency Video Coding (HEVC). VVC has been developed incrementally based on the HEVC design, with the introduction of multiple new coding tools in all building blocks of the codec architecture. This paper aims at providing an overview of VVC by highlighting the main differences compared to HEVC. It reports a compression performance analysis, based on the coding gain evaluation of each tool for various contents (including contents not used in JVET). The analysis also considers the impact of the tools in terms of encoding and decoding complexity. Global performance measures with regards to HEVC are provided in different encoding configurations and picture formats.

Introduction

The Versatile Video Coding (VVC) standardization project started by an exploratory phase in mid-2015. This phase was concluded at the end of 2017 by a Call for Proposals, and the standardization phase driven jointly by MPEG (ISO/IEC) and VCEG (ITU-T) in the Joint Video Expert Team (JVET) was launched in April 2018. After around two years development, VVC reached the Final Draft International Standard (FDIS) stage in mid-2020 [3. ]. With an estimated bitrate gain of 40% over High Efficiency Video Coding (HEVC) [1. ,2. ] for HD and 4K formats [4. ], VVC can be considered as the state-of-the-art in video compression. VVC is a hybrid video coding based on a design similar to HEVC. It has been incrementally developed by bringing enhancements to existing (HEVC) coding tools, and by adding numerous new coding tools aimed at increasing the compression performance for a variety of video contents, including Standard Dynamic Range (SDR), High Dynamic Range (HDR), 360° video and computer graphics and screen content. High-level features are also specified in the VVC core design. VVC supports layered coding, giving access to spatial, SNR and temporal scalability. VVC also introduces the concept of self-decodable sub-pictures, allowing region-wise random access, that can be used for instance for viewport dependent streaming of 360° video. VVC is therefore a versatile video coding solution which is able to address a variety of use cases and applications.

Download the paper below