Joint Spatio-Temporal Action Localization in Untrimmed Videos with Per-Frame Segmentation
Lists
Only you see this
Abstract
Inspired by the recent spatio-temporal action localization efforts with tubelets (sequences of bounding boxes), we present a new spatio-temporal action detector Segment-tube, which consists of sequences of per-frame segmentation masks. The proposed Segment-tube detector can temporally pinpoint the starting/ending frame of each action class in the presence of preceding/subsequent interference actions in untrimmed videos. Simultaneously, the Segment-tube detector produces per-frame segmentation masks instead of bounding boxes, offering superior spatial accuracy to tubelets. This is achieved by alternating iterative optimization between temporal action localization and spatial action segmentation. Experimental results on multiple datasets validate the efficacy of the proposed detector.
BibTeX
@inproceedings{duan2018joint,
author = {Xuhuan Duan and Le Wang and Changbo Zhai and Nanning Zheng and Qilin Zhang and Zhenxing Niu and Gang Hua},
title = {{Joint Spatio-Temporal Action Localization in Untrimmed Videos with Per-Frame Segmentation}},
booktitle = {ICIP},
year = {2018},
doi = {10.1109/icip.2018.8451692},
}Loading related content...