Query refinement for multimedia similarity retrieval in MARS

Advances in image processing, database management, and information retrieval has resulted in content-based multimedia retrieval to emerge as an important area of research. Typical content-based retrieval systems allow users to specify queries by providing examples of objects similar to the ones they wish to retrieve. Due to the sub-jective nature of retrieval, it is unlikely that the answers to the ‘starting query ’ will satisfy the user’s information need. Rather, among answers retrieved, the user may find one or more objects that are closer to what she has in mind compared to the original examples. In the Multimedia Analysis and Retrieval System (MARS), we have explored query refinement techniques to modify the query based on the relevance feedback of the user on the retrieved objects. Query refinement