Intelligent User Interface Agents in Content-Based Image Retrieval

The ability to browse and retrieve images based upon image content is of rapidly growing importance. This paper describes a content-based image retrieval (CBIR) system that combines dynamic, user-driven search capabilities. The system's ability to support both query-by-sketch and query-by-example is enhanced by the use of intelligent user interface agents (UIAs). The UIAs use both neural networks and an expert reasoning system to assist with relevance feedback. In addition to describing the system, a new CBIR evaluation metric is presented