A sparse support vector machine approach to region-based image categorization

Automatic image categorization using low-level features is a challenging research topic in computer vision. In this paper, we formulate the image categorization problem as a multiple-instance learning (MIL) problem by viewing an image as a bag of instances, each corresponding to a re-gion obtained from image segmentation. We propose a new solution to the resulting MIL problem. Unlike many exist-ing MIL approaches that rely on the diverse density frame-work, our approach performs an effective feature mapping through a chosen metric distance function. Thus the MIL problem becomes solvable by a regular classification algo-rithm. Sparse SVM is adopted to dramatically reduce the regions that are needed to classify images. The selected re-gions by a sparse SVM approximate to the target concepts in the traditional diverse density framework. The proposed approach is a lot more efficient in computation and less sen-sitive to the class label uncertainty. Experimental results are included to demonstrate the effectiveness and robust-ness of the proposed method. 1