典型文献
Accelerating temporal action proposal generation via high performance computing
文献摘要：
Temporal action proposal generation aims to output the starting and ending times of each potential action for long videos and often suffers from high computation cost.To addr-ess the issue,we propose a new temporal convolution network called Multipath Temporal ConvNet(MTCN).In our work,one novel high performance ring parallel architecture based is further introduced into temporal action proposal generation in order to respond to the requirements of large memory occupa-tion and a large number of videos.Remarkably,the total data transmission is reduced by adding a connection between multiple-computing load in the newly developed architecture.Compared to the traditional Parameter Server architecture,our parallel architecture has higher efficiency on temporal action detection tasks with multiple GPUs.We conduct experiments on ActivityNet-1.3 andTHUMOS 14,whereourmethodoutperforms-other state-of-art temporal action detection methods with high recall and high temporal precision.In addition,a time metric is further proposed here to evaluate the speed performancein the distributed training process.
文献关键词：
中图分类号：
作者姓名：
Tian WANG;Shiye LEI;Youyou JIANG;Choi CHANG;Hichem SNOUSSI;Guangcun SHAN;Yao FU
作者机构：
Institute of Artificial Intelligence,Beihang University,Beijing 100191,China;School of Automation Science and Electrical Engineering,Beihang University,Beijing 100191,China;School of Software,Tsinghua University,Beijing 100084,China;Department of Computer Engineering,Gachon University,Seongnam 13120,South Korea;Institute Charles Delaunay-LM2S FRE CNRS 2019,University of Technology of Troyes,Troyes 10010,France;School of Instrumentation Science and Opto-electronics Engineering,Beihang University,Beijing 100191,China;Changchun Institute of Optics,Fine Mechanics and Physics,Chinese Academy of Sciences,Changchun 130033,China
文献出处：
引用格式：
[1]Tian WANG;Shiye LEI;Youyou JIANG;Choi CHANG;Hichem SNOUSSI;Guangcun SHAN;Yao FU-.Accelerating temporal action proposal generation via high performance computing)[J].计算机科学前沿,2022(04):56-65
A类：
addr,ess,MTCN,andTHUMOS,whereourmethodoutperforms,performancein
B类：
Accelerating,temporal,action,proposal,generation,via,computing,Temporal,aims,output,starting,ending,times,each,potential,long,videos,often,suffers,from,computation,cost,To,issue,convolution,network,called,Multipath,ConvNet,In,one,novel,ring,parallel,architecture,further,introduced,into,order,respond,requirements,large,memory,occupa,number,Remarkably,total,data,transmission,reduced,by,adding,connection,between,multiple,load,newly,developed,Compared,traditional,Parameter,Server,has,higher,efficiency,detection,tasks,GPUs,We,conduct,experiments,ActivityNet,other,state,methods,recall,precision,addition,metric,proposed,evaluate,speed,distributed,training,process
AB值：
0.56107
相似文献
机标中图分类号，由域田数据科技根据网络公开资料自动分析生成，仅供学习研究参考。
