Exploiting Contextual Information for Image Re-Ranking and Rank Aggregation in Image Retrieval Tasks

• Daniel Carlos Guimarães Pedronette UNICAMP
• Ricardo da S. Torres UNICAMP

Abstract

In Content-based Image Retrieval (CBIR) systems, accurately ranking collection images is of great relevance, since users are interested in the returned images placed at the first positions. However, CBIR systems often consider only pairwise image analysis, i.e., they compute the similarity measures by taking into account only pairs of images, ignoring the rich information encoded in the relations among several images. On the other hand, the user perception usually considers the query specification and responses in a given context. This PhD work proposed five novel re-ranking and rank aggregation algorithms aiming at exploiting contextual information for improving the effectiveness of CBIR systems. We also proposed approaches for combining re-ranking and rank aggregation methods and for executing them efficiently on GPUs.

References

Jegou, H., Schmid, C., Harzallah, H., and Verbeek, J. (2010). Accurate image search using the contextual dissimilarity measure. IEEE Transactions on Pattern Analysis and Machine Intelligence, 32(1):2–11.

Kozievitch, N., Pedronette, D. C. G., da S. Torres, R., Santanchè, A., Fox, E., and Calumby, R. T. (2011). An infrastructure for searching and harvesting complex image objects. I3 Journal, 11(2):39–68.

Li, L. T., Almeida, J., Pedronette, D. C. G., Penatti, O. A. B., and da Silva Torres, R. (2012a). A multimodal approach for video geocoding. In MediaEval’2012.

Li, L. T., Pedronette, D. C. G., Almeida, J., Penatti, O. A. B., Calumby, R. T., and da Silva Torres, R. (2012b). Multimedia multimodal geocoding. In SIGSPATIAL/GIS’2012, pages 474–477.

Li, L. T., Pedronette, D. C. G., Almeida, J., Penatti, O. A. B., Calumby, R. T., and da Silva Torres, R. (2013). Accurate video multimodal geocoding. Multimedia Tools and Applications. Major Revision.

Pedronette, D. C. G. (2012). Exploiting Contextual Information for Image Re-Ranking and Rank Aggregation in Image Retrieval Tasks. PhD thesis, Campinas, SP, Brazil. Advisor: Ricardo da S. Torres.

Pedronette, D. C. G., Almeida, J., and da S. Torres, R. (2013a). A scalable re-ranking method for content-based image retrieval. Information Sciences. Major Revision.

Pedronette, D. C. G. and da S. Torres, R. (2010a). Distances correlation for re-ranking in content-based image retrieval. In Conf. on Graphics, Patterns and Images (SIBGRAPI’2010), volume 1, pages 1 – 8.

Pedronette, D. C. G. and da S. Torres, R. (2010b). Exploiting contextual information for image re-ranking. In Iberoamerican Congress on Pattern Recognition (CIARP’2010), pages 541–548.

Pedronette, D. C. G. and da S. Torres, R. (2010c). Shape retrieval using contour features and distance optmization. In International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications (VISAPP’2010), volume 1, pages 197 – 202.

Pedronette, D. C. G. and da S. Torres, R. (2011a). Exploiting clustering approaches for image re-ranking. Journal of Visual Languages and Computing, 22(6):453–466.

Pedronette, D. C. G. and da S. Torres, R. (2011b). Exploiting contextual information for rank aggregation. In International Conference on Image Processing (ICIP’2011), pages 97–100.

Pedronette, D. C. G. and da S. Torres, R. (2011c). Exploiting contextual spaces for image re-ranking and rank aggregation. In International Conference on Multimedia Retrieval (ICMR’11), pages 13:1–13:8.

Pedronette, D. C. G. and da S. Torres, R. (2011d). Image re-ranking and rank aggregation based on similarity of ranked lists. In C. Analysis of Images and Patterns (CAIP’2011), volume 6854, pages 369–376.

Pedronette, D. C. G. and da S. Torres, R. (2012a). Combining re-ranking and rank aggregation methods. In Iberoamerican Congress on Pattern Recognition (CIARP’2012), pages 170–178.

Pedronette, D. C. G. and da S. Torres, R. (2012b). Exploiting contextual information for image re-ranking and rank aggregation. International Journal of Multimedia Information Retrieval, 1(2):115–128.

Pedronette, D. C. G. and da S. Torres, R. (2012c). Exploiting pairwise recommendation and clustering strategies for image re-ranking. Information Sciences, 207(1):19–34.

Pedronette, D. C. G. and da S. Torres, R. (2013). Image re-ranking and rank aggregation based on similarity of ranked lists. Pattern Recognition. To appear, On-Line.

Pedronette, D. C. G., da S. Torres, R., Borin, E., and Breternitz, M. (2012a). Efficient image re-ranking computation on GPUs. In Int. Symposium Parallel Distributed Processing (ISPA’2012), pages 95–102.

Pedronette, D. C. G., da S. Torres, R., Borin, E., and Breternitz, M. (2012b). Efficient image re-ranking computation using GPUs. In Escola Regional de Alto Desempenho São Paulo (ERAD’2012).

Pedronette, D. C. G., da S. Torres, R., Borin, E., and Breternitz, M. (2013b). Efficient image re-ranking and rank aggregation on GPUs. Parallel Computing. Submitted.

Pedronette, D. C. G., da S. Torres, R., and Tripodi, R. C. (2013c). Using contextual spaces for image re-ranking and rank aggregation. Multimedia Tools and Applications. To appear, On-Line.