We are interested in how the human visual system transforms perception of human movement into the recognition of actions and the formation of imitative movements. We have investigated how different styles of movement, particularly emotional styles, can be recognized. To better understand the information available in human movement we have examined the effectiveness of exaggerating movement prototypes as well as used techniques of automatic pattern classification to quantify levels of human performance at recognizing movement. We have been trying to apply our basic research results to the design of interactions with humanoid robots.
