2.5D face recognition under tensor manifold metrics

Lee Ying Chong, Andrew Beng Jin Teoh, Thian Song Ong, Siew Chin Chong

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

5 Citations (Scopus)


Gabor-based region covariance matrix (GRCM) is a very flexible face descriptor where it allows different combination of features to be fused to construct a covariance matrix. GRCM resides on Tensor manifold where the computation of geodesic distance between two points requires the consideration of geometry characteristics of the manifold. Affine Invariant Riemannian Metric (AIRM) is the most widely used geodesic distance metric. However, it is computationally heavy. This paper investigates several geodesic distance metrics on Tensor manifold to find out the alternative speedy method for 2.5D face recognition using GRCM. Besides, we propose a feature-level fusion for 2.5D partial and 2D data to enhance the recognition performance.

Original languageEnglish
Title of host publicationNeural Information Processing - 21st International Conference, ICONIP 2014, Proceedings
EditorsChu Kiong Loo, Keem Siah Yap, Kok Wai Wong, Andrew Teoh, Kaizhu Huang
PublisherSpringer Verlag
Number of pages8
ISBN (Electronic)9783319126425
Publication statusPublished - 2014

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Bibliographical note

Publisher Copyright:
© Springer International Publishing Switzerland 2014.

All Science Journal Classification (ASJC) codes

  • Theoretical Computer Science
  • General Computer Science


Dive into the research topics of '2.5D face recognition under tensor manifold metrics'. Together they form a unique fingerprint.

Cite this