Deep Energies for Estimating Three-Dimensional Facial Pose and Expression

作     者：Jane Wu Michael Bao Xinwei Yao Ronald Fedkiw Jane Wu;Michael Bao;Xinwei Yao;Ronald Fedkiw

作者机构：Department of Computer ScienceStanford University353 Jane Stanford WayStanfordCA 94305USA Epic Games620 Crossroads BlvdCaryNC 27518USA 

出 版 物：《Communications on Applied Mathematics and Computation》 (应用数学与计算数学学报（英文）)

年 卷 期：2024年第6卷第2期

页      面：837-861页

核心收录：

学科分类：0809[工学-电子科学与技术（可授工学、理学学位）] 08[工学] 0714[理学-统计学（可授理学、经济学学位）] 0701[理学-数学] 0812[工学-计算机科学与技术（可授工学、理学学位）] 

基　　金：supported in part by the Office of Naval Research(ONR)N00014-13-1-0346,ONR N00014-17-1-2174,ARL AHPCRC W911NF-07-0027 generous gifts from Amazon and Toyota supported in part by the VMWare Fellowship in Honor of Ole Agesen supported in part by the Stanford School of Engineering Fellowship 

主　　题：Numerical optimization Neural networks Motion capture Face tracking 

摘      要：While much progress has been made in capturing high-quality facial performances using motion capture markers and shape-from-shading,high-end systems typically also rely on rotoscope curves hand-drawn on the *** curves are subjective and difficult to draw consistently;moreover,ad-hoc procedural methods are required for generating matching rotoscope curves on synthetic renders embedded in the optimization used to determine three-dimensional(3D)facial pose and *** propose an alternative approach whereby these curves and other keypoints are detected automatically on both the image and the synthetic renders using trained neural networks,eliminating artist subjectivity,and the ad-hoc procedures meant to mimic *** generally,we propose using machine learning networks to implicitly define deep energies which when minimized using classical optimization techniques lead to 3D facial pose and expression estimation.