Presentation 2004/12/9
Real-time Talking Head System based on Principal Component Analysis
Keisuke Kinoshita, Takaaki Kuratate,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) In this paper, we describe a system that can map one's facial motion to different person's computer graphics facial animation. The facial motion is obtained by motion capture system which measures 3D coordinates of infra-red LED markers placed on one's face. The features how a different person speaks are modeled by 3D facial shape information of pre-defined basic nine postures characteristic to speech. These information are analyzed using PCA(principal component analysis), and they are linearly mapped based on the correspondences of the nine postures. Due to PCA, the number of parameters required to generate facial animation is only nine: three coefficients of dominant principal components and six rigid head motion. The proposing system can process all these procedures in real-time on computers connected through network.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Facial Animation / Principal Component Analysis / Speech / Computer Graphics
Paper # HIP2004-71
Date of Issue

Conference Information
Committee HIP
Conference Date 2004/12/9(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Vice Chair

Paper Information
Registration To Human Information Processing (HIP)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Real-time Talking Head System based on Principal Component Analysis
Sub Title (in English)
Keyword(1) Facial Animation
Keyword(2) Principal Component Analysis
Keyword(3) Speech
Keyword(4) Computer Graphics
1st Author's Name Keisuke Kinoshita
1st Author's Affiliation ATR Human Infomation Science Laboratories()
2nd Author's Name Takaaki Kuratate
2nd Author's Affiliation ATR Human Infomation Science Laboratories
Date 2004/12/9
Paper # HIP2004-71
Volume (vol) vol.104
Number (no) 525
Page pp.pp.-
#Pages 6
Date of Issue