Output details
13 - Electrical and Electronic Engineering, Metallurgy and Materials
University of Surrey
Return to search
Output 0 of 0 in the submission
Article title
Action recognition with appearance - motion features and fast search trees
Type
D - Journal article
Title of journal
Computer Vision and Image Understanding
Article number
-
Volume number
115
Issue number
3
First page of article
426
ISSN of journal
10773142
Year of publication
2011
URL
-
Number of additional authors
-
Additional information
This paper introduces a new approach to categorise and localise objects and actions using few frames per sequence. This approach was extensively tested in application scenarios including an UAV robot designed for the MoD Grand Challenge 2008. For design, construction and control of an autonomous reconnaissance robot the Swarm Systems Team received the 'Best Innovative Idea Award' from the Ministry of Defence (UK) in 2008. This approach has been exploited in a project with the BBC to classify the Olympic Games footage. It is now being extended to behaviour analysis for early diagnosis of farm animal disease.
Interdisciplinary
-
Cross-referral requested
-
Research group
None
Proposed double-weighted
No
Double-weighted statement
-
Reserve for a double-weighted output
No
Non-English
No
English abstract
-
