An Efficient Algorithm for Content-Based Human Motion Retrieval

With the development of motion capture techniques; more and more 3D motion libraries become available. The growing amount of motion capture data requires more efficient and effective methods for indexing, searching and retrieving. In many cases, the user will only have a sketchy idea of which kind of motion to look for in the motion database. In consequence, the description about the query movement is a bottleneck for motion retrieval system. This paper presents a framework that can describe and handle the query scenes effectively. Our content-based retrieval system supports two kinds of query modes: textual query mode and query-by-example mode. By using various kinds of qualitative features and adaptive segments of motion capture data stream, our indexing and retrieval methods are carried out at the segment level rather than at the frame level, making them quite efficient. Some experimental examples are given to demonstrate the effectiveness and efficiency of proposed algorithms.