Abstract
This paper demonstrates that Affinity Propagation (AP) outperforms Kmeans for sub-topic clustering of web image retrieval. A SVM visual images retrieval system is built, and then clustering is performed on the results of each topic. Then we heighten the diversity of the 20 top results, by moving into the top the image with the lowest rank in each cluster. Using 45 dimensions Profile Entropy visual Features, we show for the 39 topics of the imageCLEF08 web image retrieval clustering campaign on 20K IAPR images, that the Cluster-Recall (CR) after AP is 13% better than the baseline without clustering, while the Precision stays almost the same. Moreover, CR and Precision without clustering are altered by Kmeans. We finally discuss that some high-level topics require text information for good CR, and that more discriminant visual features would also allow Precision enhancement after AP.
Work supported by the French National Agency of Research (ANR-06-MDCA-002) & Research Fund for the Dr. Program of Higher Education of China (200803591024). We thank P. Mulhem (LIG) for providing the training labels and Kmeans results.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
Glotin, H.: Robust Information Retrieval and Perception for a Scaled Lego-Audio-Video Multi-structuration. Pr. Habilitation Thesis, Univ. Toulon (2007)
Glotin, Zhao: LSIS TREC VIDEO 2008 High Level Feature Shot Segmentation using Compact Profil Entropy. In: NIST TRECVIDEO 2008 notebook (2008)
Tollari, Glotin: Learning Optimal Visual Features from Web Sampling in Online Image Retrieval. In: IEEE Conf. Acoustics Speech Signal Image Proc. (2008)
Glotin, Zhao: LSIS Imageclef Photo: combining text with entropic pixel features for texto-visual photo retrieval. CLEF keynotes (2008)
Thomas, A., Paul, C., Mark, S., Michael, G.: Overview of the ImageCLEFphoto 2008 Photo Retrieval Task Eval. Systems for Multilingual and Multimodal Information Access. In: 9th Wkp of the Cross-Language Eval. (2008)
Grubinger, Clough, Muller, Deselaers: The IAPR TC-12 benchmark: A new evaluation resource for visual information systems. In: Proc. OntoImage Language Resources for Content-Based Image Retrieval Wkp, with LREC (2006)
Frey, B., Dueck, D.: Clustering by Passing Messages Between Data Points. Science 315, 972–976 (2007)
Mulhem, et al.: LIG working notes on ImageCLEFphoto. CLEF keynotes (2008)
Tollari, S., Mulhem, P., Ferecatu, M., Glotin, H., Detyniecki, M., Gallinari, P., Sahbi, H., Zhao, Z.-Q.: A comparative study of diversity methods for different text and image retrieval approaches. In: Peters, C., et al. (eds.) CLEF 2008. LNCS, vol. 5706, pp. 585–592. Springer, Heidelberg (2009)
Ferecatu, M., Sahbi, H.: Bi-Modal Text and Image Retrieval with Diversity Enhancement. CLEF keynotes (2008)
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2009 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Glotin, H., Zhao, ZQ. (2009). Visual Affinity Propagation Improves Sub-topics Diversity without Loss of Precision in Web Photo Retrieval. In: Peters, C., et al. Evaluating Systems for Multilingual and Multimodal Information Access. CLEF 2008. Lecture Notes in Computer Science, vol 5706. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-04447-2_78
Download citation
DOI: https://doi.org/10.1007/978-3-642-04447-2_78
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-642-04446-5
Online ISBN: 978-3-642-04447-2
eBook Packages: Computer ScienceComputer Science (R0)