Title
Automatic Human Action Recognition in a Scene from Visual Inputs
Author
Bouma, H.
Hanckmann, P.
Marck, J.W.
de Penning, H.L.H.
den Hollander, R.J.M.
ten Hove, R.J.M.
van den Broek, S.P.
Schutte, K.
Burghouts, G.J.
Publication year
2012
Abstract
Surveillance is normally performed by humans, since it requires visual intelligence. However, it can be dangerous, especially for military operations. Therefore, unmanned visual-intelligence systems are desired. In this paper, we present a novel system that can recognize human actions. Central to the system is a break-down of high-level perceptual concepts (verbs) in simpler observable events. The system is trained on 3482 videos and evaluated on 2589 videos from DARPA, with for each video human annotations indicating the presence or absence of 48 verbs. The results show that our system reaches a good performance approaching the human average response.
Subject
Visual intelligence
Action recognition
Artificial intelligence
Retrieval
Computer vision
Defence Research
Defence, Safety and Security
Physics & Electronics ; Human
II - Intelligent Imaging ; DSS - Distributed Sensor Systems ; TPI - Training & Performance Innovations
TS - Technical Sciences ; BSS - Behavioural and Societal Sciences
To reference this document use:
http://resolver.tudelft.nl/uuid:7408d6d1-fd7c-4a46-88a4-5cd3a6b527ca
DOI
https://doi.org/10.1117/12.918
TNO identifier
455442
Publisher
SPIE, Bellingham, WA
Source
Unattended Ground, Sea, and Air Sensor Technologies and Applications XIV, 23 April 2012, Baltimore, Maryland, USA
Series
Proceedings of SPIE
Document type
conference paper