标题:Enhancing human action recognition via structural average curves analysis
作者:Zeng, Shichen; Lu, Guoliang; Yan, Peng
作者机构:[Zeng, Shichen; Lu, Guoliang; Yan, Peng] Shandong Univ, Sch Mech Engn, Natl Demonstrat Ctr Expt Mech Engn Educ,MOE, Key Lab High Efficiency & Clean Me 更多
通讯作者:Lu, GL;Lu, Guoliang
通讯作者地址:[Lu, GL]Shandong Univ, Sch Mech Engn, Natl Demonstrat Ctr Expt Mech Engn Educ,MOE, Key Lab High Efficiency & Clean Mech Manufacture, Jinan 250061, Sha 更多
来源:SIGNAL IMAGE AND VIDEO PROCESSING
出版年:2018
卷:12
期:8
页码:1551-1558
DOI:10.1007/s11760-018-1311-z
关键词:Action recognition; Limited training samples; Average sequences; Dynamic; programming
摘要:Human action recognition typically requires a large amount of training samples, which is often expensive and time-consuming to create. In this paper, we present a novel approach for enhancing human actions with a limited number of samples via structural average curves analysis. Our approach first learns average sequences from each pair of video samples for every action class and then gather them with original video samples together to form a new training set. Action modeling and recognition are proposed to be performed with the resulting new set. Our technique was evaluated on four benchmarking datasets. Our classification results are superior to those obtained with the original training sets, which suggests that the proposed method can potentially be integrated with other approaches to further improve their recognition performances.
收录类别:EI;SCOPUS;SCIE
资源类型:期刊论文
原文链接:https://www.scopus.com/inward/record.uri?eid=2-s2.0-85047661931&doi=10.1007%2fs11760-018-1311-z&partnerID=40&md5=579d6f6b661b86691503e9d83117f8c6
TOP