CVPR
2008
IEEE
15 years 26 days ago
2008
IEEE
In this paper, we present a novel approach for automatically learning a compact and yet discriminative appearance-based human action model. A video sequence is represented by a ba...