Cue integration through discriminative accumulation
2004 (English)In: PROCEEDINGS OF THE 2004 IEEE COMPUTER SOCIETY CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION, VOL 2, 2004, 578-585 p.Conference paper (Refereed)
Object recognition systems aiming to work in real world settings should use multiple cues in order to achieve robustness. We present a new cue integration scheme which extends the idea of cue accumulation to discriminative classifiers. We derive and test the scheme for Support Vector Machines (SVMs), but we also show that it is easily extendible to any large margin classifier Interestingly, in the case of one-class SVMs, the scheme can be interpreted as a new class of Mercer kernels for multiple cues. Experimental comparison with a probabilistic accumulation scheme is favorable to our method. Comparison with voting scheme shows that our method may suffer as the number of object classes increases. Based on these results, we propose a recognition algorithm consisting of a decision tree where decisions at each node are taken using our accumulation scheme. Results obtained using this new algorithm compare very favorably to accumulation (both probabilistic and discriminative) and voting scheme.
Place, publisher, year, edition, pages
2004. 578-585 p.
, PROCEEDINGS - IEEE COMPUTER SOCIETY CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION, ISSN 1063-6919 ; 2
Computer and Information Science
IdentifiersURN: urn:nbn:se:kth:diva-44319ISI: 000223605500076ScopusID: 2-s2.0-5044236501ISBN: 0-7695-2158-4OAI: oai:DiVA.org:kth-44319DiVA: diva2:451196
Conference on Computer Vision and Pattern Recognition Location: Washington, DC Date: JUN 27-JUL 02, 2004
QC 201110252011-10-252011-10-202011-11-02Bibliographically approved