Real-Time Face Pose Estimation from Single Range Images

We present a real-time algorithm to estimate the 3D pose of a previously unseen face from a single range im- age. Based on a novel shape signature to identify noses in range images, we generate candidates for their positions, and then generate and evaluate many pose hypotheses in parallel using modern graphics processing units (GPUs). We developed a novel error function that compares the in- put range image to precomputed pose images of an average face model. The algorithm is robust to large pose variations of ±90 ◦ yaw, ±45 ◦ pitch and ±30 ◦ roll rotation, facial ex- pression, partial occlusion, and works for multiple faces in the field of view. It correctly estimates 97.8% of the poses within yaw and pitch error of 15 ◦ at 55.8 fps. To evalu- ate the algorithm, we built a database of range images with large pose variations and developed a method for automatic ground truth annotation.

• Type: Conference
• Year: 2008
• Where: CVPR
• Authors: Michael D. Breitenstein, Daniel Küttel, Thibaut Weise, Luc J. Van Gool, Hanspeter Pfister
• Keywords: Face Detection, Gpu Programming, Parallel Processing, Pose Estimation, Range Processing, Real-time Computer Vision, Stereo Vision