Fast CU partition decision for H.266/VVC based on the improved DAG-SVM classifier model
- PDF / 1,284,650 Bytes
- 14 Pages / 595.276 x 790.866 pts Page_size
- 8 Downloads / 207 Views
REGULAR PAPER
Fast CU partition decision for H.266/VVC based on the improved DAG‑SVM classifier model Qiuwen Zhang1 · Yihan Wang1 · Lixun Huang1 · Bin Jiang1 · Xiao Wang1 Received: 29 May 2020 / Accepted: 15 August 2020 © Springer-Verlag GmbH Germany, part of Springer Nature 2020
Abstract One of the biggest changes in H.266/Versatile Video Coding (VVC) is introduced quad-tree with nested multi-type tree (QTMT) coding tree architecture, where the multi-type tree (MTT) structure in H.266/VVC includes binary tree (BT) and ternary tree (TT). Compared with H.265/High Efficiency Video Coding (HEVC) which only is divided by quad-tree (QT), the QTMT architecture makes the coding unit (CU) partition procedure more complexity. In this paper, we design a fast CU partition decision algorithm based on the improved Directed Acyclic Graph Support Vector Machine (DAG-SVM) model to reduce the complexity of CU partition. The video sequences are first encoded on the H.266/VVC and Test Model 4.0 (VTM 4.0), and the characteristics of the video sequences are extracted for training through the improved F-score method, where the correlation between a feature and CU partition is high. Then, the offline training is used for the improved DAG-SVM model. Finally, the trained DAG-SVM model is embedded in VTM 4.0 to early forecast the optimal CU partition modes. Simulation results indicate that the proposed method increases the time savings to 54.74% while maintaining the encoding performance. Furthermore, the proposed method exceeds the latest methods of H.266/VVC. Keywords CU partition decision · SVM · DAG-SVM · F-score · Classification · H.266 · VVC
1 Introduction The Joint Video Experts Team (JVET) is established by the Moving Picture Experts Group (MPEG) and the Video Coding Experts Group (VCEG) to be responsible for the project development of H.266/VVC [1]. As of early 2020, the JVET released the latest version test model VTM 8.0 [2]. The coding efficiency is improved by about 40% compared with the H.265/HEVC and Test Model (HM), while the subjective visual quality remains unchanged. The H.266/VVC utilizes a hybrid coding structure, the image partition has evolved from the single and fixed partition to diverse and flexible partition structure, which can more efficiently adapt to the encoding and decoding processing of high-resolution images [3]. However, the complex Communicated by Y. Zhang. * Qiuwen Zhang [email protected] 1
College of Computer and Communication Engineering, Zhengzhou University of Light Industry, Zhengzhou 450002, China
partition structure generally leads to significant complexity, which poses a higher challenge to the implementation of codecs. In addition, the H.266/VVC adds many new technologies and expands some of the previous techniques on the basis of H.265/HEVC, so the compression efficiency of H.266/VVC is significantly improved. The CU partition structure is one of the biggest changes in H.266/VVC, where the key concept is to add the MTT structure [4]. The H.265/ HEVC can only support QT structure, but
Data Loading...
