Academic
Publications
Action recognition using dynamics features

Action recognition using dynamics features,10.1109/ICRA.2011.5979900,Al Mansur,Yasushi Makihara,Yasushi Yagi

Action recognition using dynamics features  
BibTex | RIS | RefWorks Download
In this paper, we propose a method of action recognition using dynamics features based on physics model. The dynamics features are composed of torques from knee and hip joints of both legs and implicitly include the gravity, ground reaction forces, and the pose of the remaining body parts. These features are more discriminative than the kinematics features, and they result in a low dimensional representation of a human action which preserves much information of the original high dimensional pose. This low dimensional feature allows us to achieve a good classification performance even with a relatively small training data in a simple classification framework such as HMM. The effectiveness of the proposed method is demonstrated through experiments on the CMU motion capture dataset with various actions.
Cumulative Annual
View Publication
The following links allow you to view full publications. These links are maintained by other sources not affiliated with Microsoft Academic Search.