Steadiface: Real-Time Face-Centric Stabilization on Mobile Phones
We present Steadiface, a new real-time face-centric video stabilization method that simultaneously removes hand shake and keeps subject's head stable. We use a CNN to estimate the face landmarks and use them to optimize a stabilized head center. We then formulate an optimization problem to find a virtual camera pose that locates the face to the stabilized head center while retains smooth rotation and translation transitions across frames. We test the proposed method on fieldtest videos and show it stabilizes both the head motion and background. It is robust to large head pose, occlusion, facial appearance variations, and different kinds of camera motions. We show our method advances the state of art in selfie video stabilization by comparing against alternative methods. The whole process runs very efficiently on a modern mobile phone (8.1 ms/frame).
Steadiface is the result of a collaboration across several teams at Google. We thank valuable inputs from Andrei Kulik and Matsvei Zhdanovich for face landmark detector integration and GPU acceleration. We thank Yichang Shih, Donghui Han, Alan Yang and Jiayong Zhang for supports on face detector integration. We thank Zhijun He, Lida Wang and Chien-Yu Chen for supports on mobile platform implementations. We thank the teams managed by Bob Hung and Joy Hsu for data collection, video quality monitoring and feedbacks. Finally, we give special thanks to Jiyang Yu and Ravi Ramamoorthi for permissions to use their stabilized videos for comparisons and Henry Shim, Chin Man Bryan Ho and Lorraine Hsueh for video/photo usage permissions.