Relevance Optimization in Image Database Using Feature Space Preference Mapping and Particle Swarm Optimization

Two methods for retrieval relevance optimization using the user's feedback is proposed for a content-based image retrieval (CBIR) system. First, the feature space used in database image clustering for coarse classification is transferred to a preference feature spaceaccording to the user's feedback by a map generated by supervised training, thereby enabling to collect user-preferred images in the matching candidates. Second, the parameters in the fine-matching relaxation operation is optimized according to the user's evaluation of the retrieved image ranking using Particle Swarm Optimization. In the experiments, it is shown that the retrieval rankings are improved suiting the user's preference when feature space mapping and parameter optimization are used.

[1]  Toshikazu Kato,et al.  Intelligent visual interaction with image database systems-toward the multimedia personal interface , 1991 .

[2]  Alex Pentland,et al.  Photobook: tools for content-based manipulation of image databases , 1994, Other Conferences.

[3]  Alex Pentland,et al.  Photobook: Content-based manipulation of image databases , 1996, International Journal of Computer Vision.

[4]  Keisuke Kameyama,et al.  Optimal Parameter Selection in Image Similarity Evaluation Algorithms Using Particle Swarm Optimization , 2006, 2006 IEEE International Conference on Evolutionary Computation.

[5]  Teuvo Kohonen,et al.  Self-Organizing Maps , 2010 .

[6]  Marcello Pelillo,et al.  On the dynamics of relaxation labeling processes , 1994, Proceedings of 1994 IEEE International Conference on Neural Networks (ICNN'94).

[7]  Azriel Rosenfeld,et al.  Scene Labeling by Relaxation Operations , 1976, IEEE Transactions on Systems, Man, and Cybernetics.

[8]  Thierry Pun,et al.  Content-based query of image databases: inspirations from text retrieval , 2000, Pattern Recognit. Lett..

[9]  James Kennedy,et al.  Particle swarm optimization , 2002, Proceedings of ICNN'95 - International Conference on Neural Networks.

[10]  Shin'ichi Satoh,et al.  The SR-tree: an index structure for high-dimensional nearest neighbor queries , 1997, SIGMOD '97.

[11]  Keisuke Kameyama,et al.  Trademark retrieval by relaxation matching on fluency function approximated image contours , 2001, 2001 IEEE Pacific Rim Conference on Communications, Computers and Signal Processing (IEEE Cat. No.01CH37233).

[12]  Christos Faloutsos,et al.  QBIC project: querying images by content, using color, texture, and shape , 1993, Electronic Imaging.