Portrait of 3D Head Motion Recovery in Real Time
Head: Jing Xiao
Contact: Jing Xiao
Associated Lab: People Image Analysis Consortium
Last Project Publication Year: 2016

We developed a method to recover the full-motion (3 rotations and 3 translations) of the head using a cylindrical model. The robustness of the approach is achieved by a combination of three techniques. First, we use the iteratively re-weighted least squares (IRLS) technique in conjunction with the image gradient to deal with non-rigid motion and occlusion. Second, while tracking, the templates are dynamically updated to diminish the effects of self-occlusion and gradual lighting changes and keep tracking the head when most of the face is not visible. Third, because the dynamic templates may cause error accumulation, we re-register images to a reference frame when head pose is close to a reference pose. The performance of the real-time tracking program was evaluated using image sequences (both synthetic and real) for which ground truth head motion is known. The real sequences included pitch and yaw of as large as 40 and 75, respectively. The average recovery accuracy of the 3D rotations was found to be about 3.


AVI Movie, 2M4.
AVI Movie, 1M7.
Displaying 4 Publications
Master's Thesis, Tech. Report, CMU-RI-TR-16-37, Robotics Institute, Carnegie Mellon University, July, 2016
Conference Paper, Proceedings of Asian Conference on Computer Vision (ACCV '16), pp. 535 - 551, May, 2016
Conference Paper, Proceedings of International Conference on Autonomous Agents and MultiAgent Systems (AAMAS '13), pp. 1147 - 1148, May, 2013
Jeffrey Cohn, Takeo Kanade, Tsuyoshi Moriyama, Zara Ambadar, Jing Xiao, Jiang Gao and Hiroki Imamura
Tech. Report, CMU-RI-TR-02-06, Robotics Institute, Carnegie Mellon University, November, 2001

Current RI People