Exploiting Low-rank Structure for Discriminative Sub-categorization

Zheng Xu, Xue Li, Kuiyuan Yang and Thomas Goldstein

Abstract

In visual recognition, sub-categorization has been proposed to deal with large intra-class variance of samples in a category. Instead of learning a single classifier for each category, discriminant sub-categorization approaches divide a category into several sub-categories and simultaneously train classifiers for each sub-category. In this paper, we propose a novel approach for discriminative sub-categorization. Our method jointly trains the exemplar classifier for each positive sample to address the intra-variance of a category and exploits the low rank structure to preserve common information while discovering sub-categories. We formulate the problem as a convex objective function and introduce an efficient solver based on alternating direction method of multipliers.Comprehensive experiments on various datasets demonstrate the effectiveness and efficiency of the proposed method in both sub-category discovery and visual recognition.

Session

Poster 2

Files

Extended Abstract (PDF, 98K)
Paper (PDF, 250K)

DOI

10.5244/C.29.149
https://dx.doi.org/10.5244/C.29.149

Citation

Bibtex

@inproceedings{BMVC2015_149,
	title={Exploiting Low-rank Structure for Discriminative Sub-categorization},
	author={Zheng Xu and Xue Li and Kuiyuan Yang and Thomas Goldstein},
	year={2015},
	month={September},
	pages={149.1-149.12},
	articleno={149},
	numpages={12},
	booktitle={Proceedings of the British Machine Vision Conference (BMVC)},
	publisher={BMVA Press},
	editor={Xianghua Xie, Mark W. Jones, and Gary K. L. Tam},
	doi={10.5244/C.29.149},
	isbn={1-901725-53-7},
	url={https://dx.doi.org/10.5244/C.29.149}
}