Applications of image understanding in semantics-oriented multimedia information retrieval

This paper focuses on research on the development of semantics-oriented multimedia information retrieval techniques. Semantics-oriented information retrieval addresses the effectiveness of the retrieval. With the goal of significantly improving retrieval precision and query capability, image understanding technologies are applied to develop effective retrieval techniques. Specifically, three techniques can developed by combining IU technologies (i.e., face detection) with other existing or developed techniques (i.e., text indexing, conventional similarity matching and face identification) for image retrieval. These techniques are evaluated to demonstrate the significance of the semantics-oriented multimedia information retrieval.