Home >

news Help

Publication Information


Title
Japanese: 
English:VANT at TRECVID 2018 
Author
Japanese: 井上 中順, 白石 智裕, DROZD Aleksandr, 篠田 浩一, Shi-wook Lee, Alex Chichung Kot.  
English: Nakamasa Inoue, Chihiro Shiraishi, Aleksandr Drozd, Koichi Shinoda, Shi-wook Lee, Alex Chichung Kot.  
Language English 
Journal/Book name
Japanese: 
English:Proc. TRECVID workshop 
Volume, Number, Page        
Published date Nov. 13, 2018 
Publisher
Japanese: 
English: 
Conference name
Japanese: 
English:2018 TRECVID Workshop 
Conference site
Japanese: 
English:Maryland 
File
Official URL https://www-nlpir.nist.gov/projects/tv2018/tv18.workshop.notebook/tv18.papers/vant.pdf
 
Abstract We propose a system for activity detection, which utilizes the Action Tubelet (ACT) Detector to localize activities in video data. Our network is trained for all of activities in the ActEV dataset with a backbone convolutional neural network pre-trained on the ImageNet dataset. We inserted a thresholding module to the original ACT framework to adapt detector to the ActEV task, since activities in this task appear more sparsely distributed than those in the action detection task. Our result was 0.882 in mean-p miss@0.15rfa at the AD Leaderboard Evaluation.

©2007 Tokyo Institute of Technology All rights reserved.