Human Pose Estimation with Rotated Geometric Blur

We consider the problem of estimating the pose of a human figure in a single image. Our method uses an exemplarmatching framework, where a test image is matched to a database of exemplars upon which body joint positions have been marked. We find the best matching exemplar for a test image by employing a variant of an existing deformable template matching framework. A hierarchical correspondence process is developed to improve the efficiency of the existing framework. Quantitative results on the CMU MoBo dataset verify the effectiveness of our approach.

• Type: Conference
• Year: 2008
• Where: WACV
• Authors: Bo Chen, Nhan Nguyen, Greg Mori
• Keywords: Body Joint Positions, Computer Vision, Template Matching Framework, Test Image, WACV 2008