Improving Image Classification Quality Using Multi-View Learning

Abstract:

This paper proposes a novel multi-view learning framework which leverages the information contained in pseudo-labeled images to improve the prediction performance of image classification using multiple views of an image. In the training process, labeled images are first adopted to train view-specific classifiers independently using uncorrelated and sufficient views, and each view-specific classifier is then iteratively re-trained using initial labeled samples and additional pseudo-labeled samples based on a measure of confidence. In the classification process, the maximum entropy principle is utilized to assign appropriate category label to each unlabeled image using optimally trained view-specific classifiers. Experimental results on a general-purpose image database demonstrate the effectiveness and efficiency of the proposed multi-view semi-supervised scheme.

• Periodical: Advanced Materials Research (Volumes 1049-1050)
• Pages: 1475-1479
• DOI: https://doi.org/10.4028/www.scientific.net/AMR.1049-1050.1475
• Online since: October 2014
• Authors: Cheng Jian Sun*, Song Hao Zhu, Zhe Shi
• Keywords: Image Classification, Maximum Entropy, Multi-View Fusion, Multi-View Learning

References

[1] Z. Yuan, J. Sang, and C. Xu. Tag-aware image classification via Nested Deep Belief nets. Proceedings of the IEEE Conference on Multimedia and Expo, 2013: 1-6. 1.

DOI: 10.1109/icme.2013.6607503

[2] R. Wang, H. Guo, L. Davis, and Q. Dai. Covariance discriminative learning: A natural and efficient approach to image set classification. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2012: 2496-2503. 2.

DOI: 10.1109/cvpr.2012.6247965

[3] S. Zhong, Y. Liu, and Y. Liu. Bilinear deep learning for image classification. Proceedings of the ACM Conference on Multimedia, 2011: 343-352. 3.

[4] D. Mahajan and M. Slaney. Image classification using the web graph. Proceedings of the ACM Conference on Multimedia, 2010: 991-994. 6.

[5] X. Zhou, N. Cui, Z. Li, F. Liang, and T. Huang. Hierarchical Gaussianization for image classification. Proceedings of the IEEE Conference on Computer Vision, 2009: 1971-1977. 7.

DOI: 10.1109/iccv.2009.5459435

[6] R. Marée, P. Geurts, J. Piater, and L. Wehenkel. Random Subwindows for Robust Image Classification. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2005: 34-40. 10.

DOI: 10.1109/cvpr.2005.287

[7] H. Castro, L. Sucar, and E. Morales. Automatic Image Annotation Using a Semi-supervised Ensemble of Classifiers. Proceedings of the IEEE Conference on Pattern Recognition, Image Analysis and Applications, 2007: 487-495. 11.

DOI: 10.1007/978-3-540-76725-1_51

[8] W. Liu, D. Tao, J. Cheng, and Y. Tang. Multiview Hessian Discriminative Sparse Coding for Image Annotation. CoRR abs/1307. 3811 (2013). 13.

DOI: 10.1016/j.cviu.2013.03.007

[9] K. Sridharan and S. Kakade. An information theoretic framework for multiview learning. Proceedings of the Annual Conference on Learning Theory, 2008: 403-414. 15.

[10] A. Blum and T. Mitchell. Combining labeled and unlabeled data with co-training. Proceedings of the IEEE Conference on Learning Theory, 1998: 92-100. 16.

DOI: 10.1145/279943.279962

[11] M. Amini, N. Usunier, and C. Goutte. Learning from multiple partially observed views-an application to multilingual text categorization. Proceedings of the IEEE Conference on Neural Information Processing Systems, 2009: 28-36. 17.