Video event detection : from subvolume localization to spatio-temporal path search

Although sliding window-based approaches have been quite successful in detecting objects in images, it is not a trivial problem to extend them to detecting events in videos. We propose to search for spatio-temporal paths for video event detection. This new formulation can accurately detect and loca...

Full description

Saved in:
Bibliographic Details
Main Authors: Tran, Du, Yuan, Junsong, Forsyth, David
Other Authors: School of Electrical and Electronic Engineering
Format: Article
Language:English
Published: 2014
Subjects:
Online Access:https://hdl.handle.net/10356/103816
http://hdl.handle.net/10220/19322
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Nanyang Technological University
Language: English
id sg-ntu-dr.10356-103816
record_format dspace
spelling sg-ntu-dr.10356-1038162020-03-07T14:02:43Z Video event detection : from subvolume localization to spatio-temporal path search Tran, Du Yuan, Junsong Forsyth, David School of Electrical and Electronic Engineering DRNTU::Engineering::Electrical and electronic engineering Although sliding window-based approaches have been quite successful in detecting objects in images, it is not a trivial problem to extend them to detecting events in videos. We propose to search for spatio-temporal paths for video event detection. This new formulation can accurately detect and locate video events in cluttered and crowded scenes, and is robust to camera motions. It can also well handle the scale, shape, and intra-class variations of the event. Compared to event detection using spatio-temporal sliding windows, the spatio-temporal paths correspond to the event trajectories in the video space, thus can better handle events composed by moving objects. We prove that the proposed search algorithm can achieve the global optimal solution with the lowest complexity. Experiments are conducted on realistic video datasets with different event detection tasks, such as anomaly event detection, walking person detection, and running detection. Our proposed method is compatible to different types of video features or object detectors and robust to false and missed local detections. It significantly improves the overall detection and localization accuracy over the state-of-the-art methods. Accepted version 2014-05-12T02:46:39Z 2019-12-06T21:20:57Z 2014-05-12T02:46:39Z 2019-12-06T21:20:57Z 2014 2014 Journal Article Tran, D., Yuan, J., & Forsyth, D. (2014). Video Event Detection: From Subvolume Localization to Spatiotemporal Path Search. IEEE Transactions on Pattern Analysis and Machine Intelligence, 36(2), 404-416. 0162-8828 https://hdl.handle.net/10356/103816 http://hdl.handle.net/10220/19322 10.1109/TPAMI.2013.137 en IEEE transactions on pattern analysis and machine intelligence © 2014 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works. The published version is available at: [http://dx.doi.org/10.1109/TPAMI.2013.137]. 15 p. application/pdf
institution Nanyang Technological University
building NTU Library
country Singapore
collection DR-NTU
language English
topic DRNTU::Engineering::Electrical and electronic engineering
spellingShingle DRNTU::Engineering::Electrical and electronic engineering
Tran, Du
Yuan, Junsong
Forsyth, David
Video event detection : from subvolume localization to spatio-temporal path search
description Although sliding window-based approaches have been quite successful in detecting objects in images, it is not a trivial problem to extend them to detecting events in videos. We propose to search for spatio-temporal paths for video event detection. This new formulation can accurately detect and locate video events in cluttered and crowded scenes, and is robust to camera motions. It can also well handle the scale, shape, and intra-class variations of the event. Compared to event detection using spatio-temporal sliding windows, the spatio-temporal paths correspond to the event trajectories in the video space, thus can better handle events composed by moving objects. We prove that the proposed search algorithm can achieve the global optimal solution with the lowest complexity. Experiments are conducted on realistic video datasets with different event detection tasks, such as anomaly event detection, walking person detection, and running detection. Our proposed method is compatible to different types of video features or object detectors and robust to false and missed local detections. It significantly improves the overall detection and localization accuracy over the state-of-the-art methods.
author2 School of Electrical and Electronic Engineering
author_facet School of Electrical and Electronic Engineering
Tran, Du
Yuan, Junsong
Forsyth, David
format Article
author Tran, Du
Yuan, Junsong
Forsyth, David
author_sort Tran, Du
title Video event detection : from subvolume localization to spatio-temporal path search
title_short Video event detection : from subvolume localization to spatio-temporal path search
title_full Video event detection : from subvolume localization to spatio-temporal path search
title_fullStr Video event detection : from subvolume localization to spatio-temporal path search
title_full_unstemmed Video event detection : from subvolume localization to spatio-temporal path search
title_sort video event detection : from subvolume localization to spatio-temporal path search
publishDate 2014
url https://hdl.handle.net/10356/103816
http://hdl.handle.net/10220/19322
_version_ 1681049616114515968