Technical Program

Paper Detail

Session:Indexing and Retrieval
Time:Wednesday, May 19, 15:30 - 17:30
Presentation: Poster
Topic: Image and Multidimensional Signal Processing: Image and Video Indexing and Retrieval
Authors: Wei Jiang; Tsinghua University 
 Guihua Er; Tsinghua University 
 Qionghai Dai; Tsinghua University 
Abstract: Content-based image retrieval can be viewed as a classificationproblem, and the small sample size leaning difficulty makes itdifficult for most CBIR classifiers to get satisfactory performance. In this paper, using SVM classifier as the componentclassifier, the method of ensemble of classifiers is incorporatedinto the relevance feedback process to alleviate this problem from two aspects: 1. Within each feedback round, multiple parallel component classifiers are constructed, one over one feature subspace individually, and then are merged together to get an ensemble classifier. 2. During feedback rounds, boosting method is incorporated to sequentially combine the component classifiers over each feature subspace respectively, which further improves the classification result. Experiments over 5,000 images show that the proposed method can improve the retrieval performance consistently, without lost of efficiency.

Home -||- Organizing Committee -||- Technical Committee -||- Technical Program -||- Plenaries
Paper Submission -||- Special Sessions -||- ITT -||- Paper Review -||- Exhibits -||- Tutorials
Information -||- Registration -||- Travel Insurance -||- Housing -||- Workshops

©2015 Conference Management Services, Inc. -||- email: -||- Last updated Wednesday, April 07, 2004