A comparative study of color texture features for face analysis

  • Authors:
  • Seung Ho Lee;Hyungil Kim;Yong Man Ro

  • Affiliations:
  • Image and Video Systems Lab., Korea Advance Institute of Science and Technology (KAIST), Yuseong-gu, Daejeon, Republic of Korea;Image and Video Systems Lab., Korea Advance Institute of Science and Technology (KAIST), Yuseong-gu, Daejeon, Republic of Korea;Image and Video Systems Lab., Korea Advance Institute of Science and Technology (KAIST), Yuseong-gu, Daejeon, Republic of Korea

  • Venue:
  • CCIW'13 Proceedings of the 4th international conference on Computational Color Imaging
  • Year:
  • 2013

Quantified Score

Hi-index 0.00

Visualization

Abstract

Although color texture features have proven to be highly effective for face analysis, the comparisons between the color texture features have not been presented in the literature. The aim of this paper is to find the best way for combining color and texture features for face analysis. For this purpose, four different approaches (proposed for face recognition or facial expression recognition) of extracting color texture features are reviewed and compared through extensive experiments. Experimental results show that the texture feature extracted using color vector can achieve the highest recognition performances for both face recognition and facial expression recognition, among the color texture features presented in this paper.