Facial action unit recognition using recurrent neural networks

H. B. Vadapalli, H. Nyongesa, C. W.P. Omlin

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

1 Citation (Scopus)

Abstract

Facial expression recognition has been a major area of research in the field of computer vision and human computer interaction for more than ten years. Recently more emphasis has been laid on the individual muscle changes and their effect on the face, than the expressions themselves. The facial action coding system describes the appearance changes in the face at the muscular level with a set of action units. In this paper, we attempt to recognize these action units from image sequences. Feature extraction of the images is performed using Gabor filters, while classification is done using recurrent neural networks which have the ability to handle time-variant data. A set of 6 upper face action units and 5 lower face action units are recognized. An average recognition rate of 83.51% was achieved for the 6 upper face action units and 81.98% for the 5 lower face action units. The overall recognition rate of 82.75% was achieved for all the 11 action units.

Original languageEnglish
Title of host publicationProceedings of the 2009 International Conference on Image Processing, Computer Vision, and Pattern Recognition, IPCV 2009
Pages357-361
Number of pages5
Publication statusPublished - 2009
Externally publishedYes
Event2009 International Conference on Image Processing, Computer Vision, and Pattern Recognition, IPCV 2009 - Las Vegas, NV, United States
Duration: 13 Jul 200916 Jul 2009

Publication series

NameProceedings of the 2009 International Conference on Image Processing, Computer Vision, and Pattern Recognition, IPCV 2009
Volume1

Conference

Conference2009 International Conference on Image Processing, Computer Vision, and Pattern Recognition, IPCV 2009
Country/TerritoryUnited States
CityLas Vegas, NV
Period13/07/0916/07/09

Keywords

  • Facial action unit recognition
  • Feature extraction
  • Gabor filters
  • Recurrent neural networks

ASJC Scopus subject areas

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

Fingerprint

Dive into the research topics of 'Facial action unit recognition using recurrent neural networks'. Together they form a unique fingerprint.

Cite this