Automatic pose estimation system for human faces based on Bunch graph matching technology

被引：16

作者：

Elagin, E ^{[1
]}

Steffens, J ^{[1
]}

Neven, H ^{[1
]}

机构：

[1] Eyemat Interfaces Inc, Santa Monica, CA 90403 USA

来源：

AUTOMATIC FACE AND GESTURE RECOGNITION - THIRD IEEE INTERNATIONAL CONFERENCE PROCEEDINGS | 1998年

关键词：

D O I：

10.1109/AFGR.1998.670938

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

We present an automatic module that can determine the pose of a human face from a digitized portrait-style image. The module is integrated into a larger system called PersonSpotter, which is able to recognize people by their facts coming from a live video stream of data. The Pose Estimation Module is based on Bunch Graph :Matching and can distinguish between five different degrees of rotation in depth. The system features close to real-time performance, considerable decrease in data site and increase in the accuracy of pose recognition compared to similar systems developed in the past. Pose estimation success rate of 98.5% has been reached for a set of 210 faces rotated in various degrees and directions.

引用

页码：136 / 141

页数：2