Output details
11 - Computer Science and Informatics
University of Edinburgh
Output title
Active Matching
Type
E - Conference contribution
Name of conference/published proceedings
Computer Vision – ECCV 2008 : 10th European Conference on Computer Vision, Marseille, France, October 12-18, 2008, Proceedings, Part I
Volume number
-
Issue number
-
First page of article
72
ISSN of proceedings
-
Year of publication
2008
Number of additional authors
1
Additional information
<23> Originality: In contrast to the so-far popular brute force randomized approaches, this paper presents a novel
method to take all probabilistic priors into account in the frame-to-frame matching problem in sequential tracking.
Significance: Opened up a whole new perspective of approaching feature matching: rather than using ad-hoc thresholds, rely on information theory to guide the matching process for most effective and efficient search for matches. Led to many subsequent publications based on this idea.
Rigour: Presented at a top-tier double-blind international conference with a 27.9% acceptance rate.
Interdisciplinary
-
Cross-referral requested
-
Research group
E - Institute of Perception, Action & Behaviour
Citation count
4
Proposed double-weighted
No
Double-weighted statement
-
Reserve for a double-weighted output
No
Non-English
No
English abstract
-