Pose robust 3D face recognition using the RBFN feature

Ukil Yang, Kwanghoon Sohn

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

This paper describes a novel global shape (GS) feature of three-dimensional (3D) face data based on the Radial Basis Function Network (RBFN) as well as an extraction method of the proposed feature for 3D face recognition. The features are extracted from facial profiles based on the RBFN. To validate the robustness of the RBFN feature for pose variations, we perform experiments using the test images which consist of five pose variations, and we compare the performance of the proposed feature with those of 3D Principal Component Analysis (3D PCA) and Extended Gaussian Image (EGI). We also perform an experiment about a problem of the holes caused by occlusion region which may appear after the pose compensation of 3D data having one view point. Through these experiments, it is obvious that the RBFN feature outperforms the 3D PCA and the EGI for 3D facial recognition under the pose variable environments.

Original languageEnglish
Title of host publicationProceedings of the 7th IASTED International Conference on Visualization, Imaging, and Image Processing, VIIP 2007
Pages235-240
Number of pages6
Publication statusPublished - 2007
Event7th IASTED International Conference on Visualization, Imaging, and Image Processing, VIIP 2007 - Palma de Mallorca, Spain
Duration: 2007 Aug 292007 Aug 31

Publication series

NameProceedings of the 7th IASTED International Conference on Visualization, Imaging, and Image Processing, VIIP 2007

Other

Other7th IASTED International Conference on Visualization, Imaging, and Image Processing, VIIP 2007
Country/TerritorySpain
CityPalma de Mallorca
Period07/8/2907/8/31

All Science Journal Classification (ASJC) codes

  • Computer Graphics and Computer-Aided Design
  • Computer Vision and Pattern Recognition

Fingerprint

Dive into the research topics of 'Pose robust 3D face recognition using the RBFN feature'. Together they form a unique fingerprint.

Cite this