Image Retrieval Based on An Artificial Emotion Model

摘要

In the design of multimedia database system, one of the important issues is how to deal with emotion of human beings. An image retrieval based on an artificial emotion model has been proposed in this paper. The objects of the emotion model are adjectives such as romantic or elegant expressed by people who are retrieving images. Factor analysis is applied to construct an orthogonal psychological space based on sense experiments. After that, a radial basis function neural network is used for learning and memorizing the user's emotion. Then each image is automatically evaluated and annotated with adjective words in the orthogonal psychological space. Thus an emotional image retrieval system is obtained. Last an interesting experimental result is presented.