关键词: 人工智能, 图像分类, 多示例学习, 密度聚类, 支持向量机

Abstract: In order to narrow the semantic gap between low-level visual features and high-level semantic concepts in image categorization, this paper exploits the clustering information from a density clustering algorithm and the characteristics of multi-instance learning framework in distinguishing ambiguous object, proposes an image categorization method using density clustering on region features and multi-instance learning, termed as DCRF-MIL. First, it divides each image into a number of regions, re-lines up all regions into a collection, and then uses a density clustering algorithm to learn the potential distribution information of region features in the collection. Second, it treats image as bag and regions as instances. Based on the cluster distribution information of region features, the bag is mapped into a vector in the cluster distribution space. Finally, a support vector machine classifier is constructed to predict the class label of the unlabeled image. The experiments on the Corel image data set and MUSK molecular activity prediction data set show DCRF-MIL algorithm has high classification accuracy and its parameters selection are easy to select.

Key words: artificial Intelligence, image categorization, multi-instance learning, density clustering, support vector machine