An Experiment on Generic Image Classification Using Web Images

In this paper, we describe an experiment on generic image classification using a large number of images gathered from the Web as learning images. The processing consists of three steps. In the gathering stage, a system gathers images related to given class keywords from the Web automatically. In the learning stage, it extracts image features from gathered images and associates them with each class. In the classification stage, the system classifies a test image into one of classes corresponding to the class keywords by using the association between image features and classes. In the experiments, we achieved a classification rate 44.6% for generic images by using images gathered from the World-Wide Web automatically as learning images.