A Face Replacement System Based on 3D Face Model

摘要：

　　In this paper,an efficient face replacement system based on 3D model is developed,which consists of four parts.First,given an input video containing a source face rotating from frontal to profile,after two manual clicks on a profile to tell the system where the eye corner and bottom of the chin are,a realistic looking source face 3D model is generated.Secondly,Random Forests Algorithm(RFA)is used to estimate target face pose.Thirdly,a source face image is rendered based on the result of pose estimation and color transfer method.Finally,Poisson image fusion algorithm for boundary blurring is used to seamlessly synthesize the source face into the target image.The effectiveness of our system is demonstrated with a set of experiments.

关键词： Face replacement Face modeling Pose estimation Color transfer Poisson image fusion

作者: Hong Song Jie Lv He Liu Qingjie Zhao

作者单位: School of Software,Beijing Institute of Technology,Beijing 100081,China School of Computer Science,Beijing Institute of Technology,Beijing 100081,China

会议类型: 国际会议

会议名称: The 2015 Chinese Intelligent Automation Conference(2015中国智能自动化会议)

会议地点: 福州

会议语种:英文

页码: 237-246

在线出版日期: 2015-05-08（万方平台首次上网日期，不代表论文的发表时间）