Graphics enhanced version of this site
Temporal Segmentation of Human Motion
Heads: Jessica K Hodgins and Fernando De la Torre Frade
Contact: Fernando De la Torre Frade (ftorre@cs.cmu.edu)
Mailing address:
Carnegie Mellon University
Robotics Institute
5000 Forbes Avenue
Pittsburgh, PA 15213
Temporal segmentation of human motion into different activities is a crucial step for understanding and building computational models of human activity. This project explores robust methods to temporally segment, into coherent temporal patterns, streams of human behavior coming from motion capture data of several subjects. Several issues contribute to the challenge of this task. These include large variability in the temporal scale and periodicity of human actions, inter-personal differences in motion patterns, and the exponential nature of all possible action combinations.