On the Dynamic Pattern Analysis and Recognition
- Speaker: Dr Ke Chen (University of Manchester)
- Host: Gavin Brown
- 8th November 2006 at 14:15 in 1.5
In the real world, most of objects can be treated as one kind of patterns in general. Although patterns are ubiquitous and of miscellaneous forms, they can be generally divided into two categories: static and dynamic patterns. The nature of dynamic patterns results in many challenging real world problems and therefore the dynamic pattern analysis and recognition become a focus of the current pattern recognition research. In this seminar we discuss the nature of dynamic patterns and useful strategies for tackle problems in the dynamic pattern analysis and recognition including the representations of dynamic patterns, the use of divide-and-conquer principle and the exploitation of intrinsic contextual information, as exemplified by our previous work in machine perception. Future work in dynamic pattern analysis and recognition is also discussed.