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

Author

Elagin, Egor ; Steffens, Johannes ; Neven, Hartmut

Author_Institution

Lab. for Comput. & Biol. Vision, Univ. of Southern California, Los Angeles, CA, USA

fYear

1998

fDate

14-16 Apr 1998

Firstpage

136

Lastpage

141

Abstract

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 faces 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 size 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

Keywords

face recognition; image matching; PersonSpotter; automatic pose estimation system; bunch graph matching technology; digitized portrait-style image; human faces; live video stream; Biology computing; Computer interfaces; Computer vision; Face detection; Face recognition; Head; Humans; Image recognition; Laboratories; Switches;

fLanguage

English

Publisher

ieee

Conference_Titel

Automatic Face and Gesture Recognition, 1998. Proceedings. Third IEEE International Conference on

Conference_Location

Nara

Print_ISBN

0-8186-8344-9

Type

conf

DOI

10.1109/AFGR.1998.670938

Filename

670938