Machine Learning

For realizing automatic analysis and understanding of multimedia information content, we research the effective methods of machine learning and information content representation, as well as the related mechanism in human perception. Our work targets at multimedia information retrieval applications and is related with the corresponding research in neural science, brain science, cognitive psychology and etc. The current main problems we are interested in include similarity measure, image content representation, classifier learning, image clustering, image data mining, web data mining, and etc.

Mobile Search

We research the methods and techniques of multimedia information retrieval on mobile computation platform. Our purpose is to make people using mobile computing equipments find out the mobile information that they want, conveniently, anytime, and anywhere. The current main problems we are working on include interactive image segmentation, video based image segmentation, image indexing, image feature extraction on compression domain, mobile retrieval system, and etc.

Medical Imaging Analysis

We research medical image retrieval methods and techniques for the applications of computer aided detection and diagnosis of diseases. Our purpose is to improve the accuracy of disease detection and diagnosis. The current main problems that we are interested include imaging sign detection, imaging sign classification, imaging sign database, and medical image retrieval system based on imaging signs.

Video Retrieval

We research video retrieval methods and techniques for the applications on video surveillance network. Our purpose is to make people obtain the objects that they are interested in from massive video data conveniently and efficiently. The current main problems that we are trying to solve include shot segmentation, key frame extraction, video indexing, object based video retrieval system, and etc.