Publication View

2004 IEEE International Conference on Multimedia and Expo (ICME) Online Feature Selection Based on Generalized Feature Contrast Model* (2008)

Abstract
To really bridge the gap between high-level semantics and low-levelfeatures in content-based image retrieval (CBIR), a problem that must be solved is: which features are suit-able for explaining the current query concept. In thispopec we propose a novel feature selection criterion based on a psychological similarity measurement- generalized feature contrast model, and implement an online feature selection algorithm in a boosting manner to select the most repre-sentative features and do classijication during each feed-back round. The advantage of the proposed method is: it doesn’t require Gaussian assumption for “relevant ” images as other online FS methods; it accounts for the intrinsic asymmetry between “relevant ” and “irrelevant ” image sets in CBIR online Ieaming; it is very fast. Extensive expen-ment5 have shown our algorithm’s effectiveness. 1

Publication details
Download http://citeseerx.ist.psu.edu/viewdoc/summary?doi=?doi=10.1.1.118.5787
Source http://www.cs.columbia.edu/~jwgu/icme.pdf
Contributors CiteSeerX
Repository CiteSeerX - Scientific Literature Digital Library and Search Engine (United States)
Type text
Language English
Relation 10.1.1.29.3868, 10.1.1.43.6406