Green Video Compression for Metaverse: Lessons Learned from VP9 and HEVC

Unprecedented growth in video consumption applications has been witnessed over the past decade, and new heights are now being reached with the emergence of the metaverse. As the metaverse expands, a substantial increase in the volume of digital data and the computational load on networks, data centers, and user devices is brought along. This continuous processing of information leads to significant energy consumption, resulting in a staggering amount of CO2 emissions annually. To address this challenge, the development of lightweight video compression algorithms that can effectively reduce file sizes and facilitate efficient transmission over the Internet needs to be prioritized. However, the desired level of efficiency cannot be achieved by current standards. In this study, a comprehensive analysis of the most resource-intensive task, Motion Estimation (ME), is conducted in two state-of-the-art compression algorithms used for metaverse videos: VP9 and HEVC. An implementation in Matlab that centers on the ME process of both algorithms has been developed for an exhaustive performance evaluation, allowing for an objective comparison. Furthermore, novel approaches were incorporated into the code to assess sustainability factors. The insights gained from this analysis shed light on key areas that require improvement in future video compression algorithms, paving the way for sustainable and optimized video storage and transmission in the metaverse.

Published

2023-10

Content type

Original Research

Keywords

HEVC, metaverse, motion estimation, sustainability, VP9

DOI

10.5594/M002002

ISBN

978-1-61482-964-5