3D Human Pose Estimation = 2D Pose Estimation + Matching

Speaker: Mr. Chinghang Chen
Date: 2018-03-26 (Mon) 11:00 - 12:00
Location: Auditorium 122 at CITI Building
Host: Yen-Yu Lin

Abstract:

We explore 3D human pose estimation from a single RGB image. While many approaches try to directly predict 3D pose from image measurements, we explore a simple architecture that reasons through intermediate 2D pose predictions. Our approach is based on two key observations (1) Deep neural nets have revolutionized 2D pose estimation, producing accurate 2D predictions even for poses with self occlusions. (2) Big-data sets of 3D mocap data are now readily available, making it tempting to lift predicted 2D poses to 3D through simple memorization (e.g., nearest neighbors). The resulting architecture is trivial to implement with off-the- shelf 2D pose estimation systems and 3D mocap libraries. Importantly, we demonstrate that such methods outperform almost all state-of- the-art 3D pose estimation systems, most of which directly try to regress 3D pose from 2D measurements.

Bio:

Chinghang Chen received his MSCV (master in computer vision) degree at Carnegie Mellon University in Dec 2016. Before that, he was a research assistant in Academia Sinica (June 2014 to July 2017). He is now an applied scientist at Computer Vision algorithm team, Amazon Lab126. His research interest includes object detection, human pose estimation, action recognition and scene geometry understanding.