Learning From Coding Features: High Efficiency Rate Control for AOMedia Video 1

Yi CHEN, Yunhao MAO, Shiqi WANG, Xianguo ZHANG, Sam KWONG

Research output: Journal PublicationsJournal Article (refereed)peer-review

1 Citation (Scopus)

Abstract

Rate control, which typically includes bit allocation and quantization parameter (QP) determination, plays an important role in real-world video coding applications. In this article, we propose a novel rate control scheme for AOMedia Video 1 (AV1) that provides adaptive bit allocation and effective QP determination. In particular, two supporting vector regression models are learned for the hierarchical bit allocation and frame-level parameter estimation. To train the models, the multipass coding strategy is utilized for training data acquisition. Compared to the default scheme in AV1 and the state-of-the-art method, the proposed rate control scheme achieves superior performance in terms of bitrate accuracy and coding efficiency.

Original languageEnglish
Pages (from-to)16-25
Number of pages10
JournalIEEE Multimedia
Volume30
Issue number4
Early online date10 Apr 2023
DOIs
Publication statusPublished - Dec 2023
Externally publishedYes

Bibliographical note

Publisher Copyright:
© 1994-2012 IEEE.

Funding

The authors would like to thank the associate editor and anonymous reviewers for their valuable comments, which significantly helped us improve the presentation of this article. This work was supported in part by the Shenzhen Science and Technology Program (Project JCYJ20220530140816037), in part by National Natural Science Foundation of China under Grant 62022002, in part by the Hong Kong Innovation and Technology Commission (InnoHK Project Centre for Intelligent Multidimensional Data Analysis), in part by the Hong Kong General Research Grant-Research Grant Council General Research Fund under Grant 11209819 (City University of Hong Kong 9042816) and Grant 11203820 (9042598), and in part by the Tencent Rhinoceros Bird project.

Keywords

  • Adaptation models
  • Bit rate
  • Encoding
  • Machine learning
  • Parameter estimation
  • Training
  • Video coding

Fingerprint

Dive into the research topics of 'Learning From Coding Features: High Efficiency Rate Control for AOMedia Video 1'. Together they form a unique fingerprint.

Cite this