Temporal Action Localization in Untrimmed Videos
Date : 25 January 2018 (Thursday)
Time : 2:30 pm – 4:30 pm
Venue : M08
Ms. LI RUIMIN
Research Assistant, Department of Computer Science
Chu Hai College of Higher Education
Temporal action localization: given a long untrimmed video, when does a specific action start and end? This is a challenging topic because algorithms must reason not only about whether an action occurs somewhere in a video, but also on the temporal extent of when it occurs. This workshop will give an introduction of this topic. In addition, one representative paper (2016 CVPR) entitled “End-to-end Learning of Action Detection from Frame Glimpses in Videos” will be shared. This paper proposed a fully end-to-end approach for action detection in videos that learns to directly predict the temporal bounds of actions. The principle and content of this method will be explained in detail.