A practical SVM-based algorithm for ordinal regression in image retrieval

Most current learning algorithms for image retrieval are based on dichotomy relevance judgement (relevant and non-relevant), though this measurement of relevance is too coarse. To better identify the user needs and preference, a good retrieval system should be able to handle multilevel relevance judgement. In this paper, we focus on relevance feedback with multilevel relevance judgment, where the relevance feedback is considered as an ordinal regression problem. Herbrich has proposed a support vector learning algorithm for ordinal regression based on the Linear Utility Model. His algorithm is intrinsically to train a SVM on a new derived training set, whose size increases rapidly when the original training set gets bigger. This property limits its applicability in relevance feedback, due to real-time requirement of the interactive process. By thoroughly analyzing Herbrich's algorithm, we first propose a new model for ordinal regression, called Cascade Linear Utility Model, then a practical SVM-based algorithm for image retrieval upon it. Our new algorithm is tested on a real-world image database, and compared with other three algorithms capable to handle multilevel relevance judgment. The experimental results show that the retrieval performance of our algorithm is comparable with that of Herbrich's algorithm but with only a fraction of its computational time, and apparently outperform the other methods.

[1]  Thomas S. Huang,et al.  One-class SVM for learning in image retrieval , 2001, Proceedings 2001 International Conference on Image Processing (Cat. No.01CH37205).

[2]  Paul B. Kantor,et al.  A study of information seeking and retrieving. I. background and methodology , 1988 .

[3]  Thomas S. Huang,et al.  Relevance feedback: a power tool for interactive content-based image retrieval , 1998, IEEE Trans. Circuits Syst. Video Technol..

[4]  Thomas S. Huang,et al.  Optimizing learning in image retrieval , 2000, Proceedings IEEE Conference on Computer Vision and Pattern Recognition. CVPR 2000 (Cat. No.PR00662).

[5]  Nello Cristianini,et al.  Large Margin DAGs for Multiclass Classification , 1999, NIPS.

[6]  Edward Y. Chang,et al.  Support vector machine active learning for image retrieval , 2001, MULTIMEDIA '01.

[7]  Klaus Obermayer,et al.  Support vector learning for ordinal regression , 1999 .

[8]  Thomas G. Dietterich,et al.  Solving Multiclass Learning Problems via Error-Correcting Output Codes , 1994, J. Artif. Intell. Res..

[9]  Klaus Obermayer,et al.  Regression Models for Ordinal Data: A Machine Learning Approach , 1999 .

[10]  Amanda Spink,et al.  Examining Different Regions of Relevance: From Highly Relevant to Not Relevant , 1998 .

[11]  Chahab Nastar,et al.  Relevance feedback and category search in image databases , 1999, Proceedings IEEE International Conference on Multimedia Computing and Systems.

[12]  K. Obermayer,et al.  Learning Preference Relations for Information Retrieval , 1998 .

[13]  Bo Zhang,et al.  Support vector machine learning for image retrieval , 2001, Proceedings 2001 International Conference on Image Processing (Cat. No.01CH37205).

[14]  Qi Tian,et al.  Discriminant-EM algorithm with application to image retrieval , 2000, Proceedings IEEE Conference on Computer Vision and Pattern Recognition. CVPR 2000 (Cat. No.PR00662).

[15]  Vladimir N. Vapnik,et al.  The Nature of Statistical Learning Theory , 2000, Statistics for Engineering and Information Science.

[16]  Vladimir Vapnik,et al.  Statistical learning theory , 1998 .

[17]  Yiyu Yao,et al.  Measuring Retrieval Effectiveness Based on User Preference of Documents , 1995, J. Am. Soc. Inf. Sci..

[18]  Paul A. Viola,et al.  Boosting Image Retrieval , 2004, International Journal of Computer Vision.

[19]  Yiyu Yao.  Measuring retrieval effectiveness based on user preference of documents , 1995 .

[20]  Christos Faloutsos,et al.  MindReader: Querying Databases Through Multiple Examples , 1998, VLDB.

[21]  Michael B. Eisenberg.  Measuring relevance judgments , 1988, Inf. Process. Manag..

[22]  Hanqing Lu,et al.  Multilevel Relevance Judgement, Loss Function, and Performance Measure in Image Retrieval , 2003, CIVR.

[23]  Thorsten Joachims,et al.  Optimizing search engines using clickthrough data , 2002, KDD.

[24]  S. K. Michael Wong,et al.  Linear structure in information retrieval , 1988, SIGIR '88.

[25]  John C. Platt,et al.  Fast training of support vector machines using sequential minimal optimization, advances in kernel methods , 1999 .

[26]  Tom Minka,et al.  Modeling user subjectivity in image libraries , 1996, Proceedings of 3rd IEEE International Conference on Image Processing.

[27]  SpinkAmanda,et al.  From highly relevant to not relevant , 1998 .

[28]  Edward Y. Chang,et al.  SVM binary classifier ensembles for image classification , 2001, CIKM '01.

[29]  M. E. Maron,et al.  On Relevance, Probabilistic Indexing and Information Retrieval , 1960, JACM.

[30]  Paul A. Viola,et al.  Boosting Image Retrieval , 2000, Proceedings IEEE Conference on Computer Vision and Pattern Recognition. CVPR 2000 (Cat. No.PR00662).