Weakly Supervised Action Localization by Sparse Temporal Pooling Network

CVPR 2018 Phuc NguyenTing LiuGautam PrasadBohyung Han

We propose a weakly supervised temporal action localization algorithm on untrimmed videos using convolutional neural networks. Our algorithm learns from video-level class labels and predicts temporal intervals of human actions with no requirement of temporal localization annotations... (read more)

PDF Abstract

Evaluation results from the paper

Task Dataset Model Metric name Metric value Global rank Compare
Weakly Supervised Action Localization THUMOS 2014 STPN MAP 16.2 # 3