Real-Time Facial Pose Estimation and Tracking by Coarse-to-Fine Iterative Optimization

作     者：Xiaolong Yang Xiaohong Jia Mengke Yuan Dong-Ming Yan Xiaolong Yang;Xiaohong Jia;Mengke Yuan;Dong-Ming Yan

作者机构：the Key Laboratory of Mathematics MechanizationAcademy of Mathematics and Systems SciencesChinese Academy of SciencesBeijing 100190China the University of Chinese Academy of SciencesBeijing 100049China the National Laboratory of Pattern RecognitionInstitute of AutomationChinese Academy of SciencesBeijing 100190China 

出 版 物：《Tsinghua Science and Technology》 (清华大学学报（自然科学版（英文版）)

年 卷 期：2020年第25卷第5期

页      面：690-700页

核心收录：

学科分类：08[工学] 080203[工学-机械设计及理论] 0802[工学-机械工程] 

基　　金：supported by the National Natural Science Foundation of China(Nos.61872354,61772523,61620106003,and 61802406) the National Key R&D Program of China(No.2019YFB2204104) the Beijing Natural Science Foundation(Nos.L182059 and Z190004) the Intelligent Science and Technology Advanced Subject Project of University of Chinese Academy of Sciences(No.115200S001) the Alibaba Group through Alibaba Innovative Research Program 

主　　题：facial pose recognition facial pose estimation real-time tracking 

摘      要：We present a novel and efficient method for real-time multiple facial poses estimation and tracking in a single frame or ***,we combine two standard convolutional neural network models for face detection and mean shape learning to generate initial estimations of alignment and ***,we design a bi-objective optimization strategy to iteratively refine the obtained *** strategy achieves faster speed and more accurate ***,we further apply algebraic filtering processing,including Gaussian filter for background removal and extended Kalman filter for target prediction,to maintain real-time tracking *** general RGB photos or videos are required,which are captured by a commodity monocular camera without any priori or *** demonstrate the advantages of our approach by comparing it with the most recent work in terms of performance and accuracy.