Automatic Image Annotation Using Semantic Subspace Graph Spectral Clustering Algorithm

Abstract:

Due to the existing of the semantic gap, images with the same or similar low level features are possibly different on semantic level. How to find the underlying relationship between the high-level semantic and low level features is one of the difficult problems for image annotation. In this paper, a new image annotation method based on graph spectral clustering with the consistency of semantics is proposed with detailed analysis on the advantages and disadvantages of the existed image annotation methods. The proposed method firstly cluster image into several semantic classes by semantic similarity measurement in the semantic subspace. Within each semantic class, images are re-clustered with visual features of region Then, the joint probability distribution of blobs and words was modeled by using Multiple-Bernoulli Relevance Model. We can annotate a unannotated image by using the joint distribution. Experimental results show the the effectiveness of the proposed approach in terms of quality of the image annotation. the consistency of high-level semantics and low level features is efficiently achieved.

• Periodical: Advanced Materials Research (Volumes 271-273)
• Pages: 1090-1095
• DOI: https://doi.org/10.4028/www.scientific.net/AMR.271-273.1090
• Online since: July 2011
• Authors: Yu Tang Guo, Chang Gang Han
• Keywords: Graph Spectral Clustering, Image Annotation, K-Harmonic Means, Semantic Gap

References

[1] Y. Liu, D.S. Zhang, G.J. Lu, et al: A survery of content-based image retrieval with high-level semantics. Pattern Recognition. Vol. 40( 2007), pp.262-282.

DOI: 10.1016/j.patcog.2006.04.045

[2] Y. Mori,H. Takahashi, R. Oka: Image-to-word transformation based on dividing and vector quantizing images with words. On http: /citeseer. ist. psu. edu/368129. html.

[3] P . Duygulu, K. Barnard, J.F. Freitas, et al:. Object recognition as machine translation: learning a lexicon for a fixed image vocabulary. Leture Noyes in Computer Science. Heidelberg: Springer. Vol. 23(2002), pp.97-112.

DOI: 10.1007/3-540-47979-1_7

[4] K. Barnard, P. Duygulu, D. Forsyth, et al: Mathing words and pictures. Journal of Machine Learning Research. Vol. 3(2003), pp.1107-1135.

[5] J. Jeon, V. Lavrenko, R. Mnmatha: Automatic image annotation and retrieval using cross-media relevance models. Proceedings. of the 26th Annual Intelnational ACM SIGIR Conference on Research and Development in information Retrieval, Toronto. (2003).

DOI: 10.1145/860435.860459

[6] S. L. Feng, R . Manmatha, V. Lavrenko: Multiple Bernoulli relevance models for image and video annotation. Proc. Of IEEE Computer Society Conference on Computer Vision and Pattern Recognition, Washington D C, (2004), pp.1002-1009.

DOI: 10.1109/cvpr.2004.1315274

[7] D. Stan, I. K. Sethi: Mapping low-level image features to semantic concepts. Proceedings of the Storage and Retrieval for Media Databases. (2001), pp.172-179.

DOI: 10.1117/12.410925

[8] W. Jin, R. Shi, T.S. Chua: A semi-nave bayesian method incorporating clustering with pair-wise constraints for auto image annotation. Proceedings of the ACM Multimedia. (2004).

DOI: 10.1145/1027527.1027605

[9] C. Fowlkes, S. Belongie, F. Chung, et al: Spectral grouping using the nystrom method. IEEE Transactions on Pattern Analysis and Machine Intelligence. Vol. 26(2007), pp.217-225.

DOI: 10.1109/tpami.2004.1262185

[10] B. Zhang,M. Hsu, U. Dayal: K-Harmonci Means-A spatial clustering algorithm with boosting. Proceddings of the First International Workshop On Temporal, Spatial, and Spatio-Temporal Data Mining- Revised Papers. (2000), pp.31-45.

DOI: 10.1007/3-540-45244-3_4

[11] S. Kullback, R.A. Leibler: On information and sufficiency. The Annals of Mathematical Statistics. Vol. 22(1951), pp.79-86.

DOI: 10.1214/aoms/1177729694

[12] J.B. Shi, J. Malik: Normalized cuts and image segmentation. IEEE Transactions on Pattern Analysis and Machine Intelligence. Vol. 22(2000), pp.888-905.

DOI: 10.1109/34.868688