Machine learning-based fast intra coding unit depth decision for high efficiency video coding

NCUIR > College of Electrical Engineering & Computer Science > Department of Communication Engineering > journal & Dissertation > Item 987654321/106006

Please use this identifier to cite or link to this item: https://ir.lib.ncu.edu.tw/handle/987654321/106006

Title:	Machine learning-based fast intra coding unit depth decision for high efficiency video coding
Authors:	張寶基;CHEN), 陳宗毅(ZONG-YI;FANG), 方俊才(JIUNN-TSAIR;LIU), 劉宴均(YEN-CHUN;CHANG), 張寶基(PAO-CHI
Contributors:	資訊電機學院通訊工程學系
Keywords:	Algorithms;Artificial neural networks;Coding;Learning theory;Mathematical models;Support vector machines;Training;Video compression
Date:	2016-09-01
Issue Date:	2026-04-23 13:03:32 (UTC+8)
Publisher:	Institute of Information Science;社團法人中華民國計算語言學學會
Abstract:	摘要： This paper proposes a fast coding unit (CU) depth decision algorithm for intra coding of high efficiency video coding using an artificial neural network (ANN) and a support vector machine (SVM). Machine learning provides a systematic approach for developing a fast algorithm for early CU splitting or termination to reduce intra coding computational complexity. Appropriate features for training SVM models were extracted from spatial and pixel domains of the current CU. These features were classified into three types for three SVM training models at each depth, and different weights were assigned on the basis of the ANN analysis. Experimental results showed that the proposed fast algorithm saves at most 48.5% and on average 33% encoding time with a 1.55% Bjøntegaard delta bit rate (BDBR) loss compared with HM 15.0. 出版者：社團法人中華民國計算語言學學會出版日期： 2016-09-01 出處： Journal of Information Science and Engineering, 2016-09, Vol.32 (5), p.1289-1299 資源來源：中文電子期刊服務 CEPS: Chinese Electronic Periodical Services 識別號： ISSN: 1016-2364
Appears in Collections:	[Department of Communication Engineering] journal & Dissertation

Files in This Item:

File	Description	Size	Format
index.html		0Kb	HTML	25	View/Open

社群 sharing

Loading...