Automatic facial expression recognition using modified wavelet-based salient points and gabor-wavelet filters

Nooshin Nabizadeh, Nigel John

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

1 Scopus citations

Abstract

In this paper, we present an automated approach for recognizing seven facial expressions including the neutral expression. The approach is based upon efficient feature extraction, feature compression, and an artificial neural network (ANN) classification. In the proposed method, the basic components of face, eyes, eyebrow, and mouth, are first segmented from the whole face using modified Wavelet based salient points. Then, the features of the eye and the mouth are extracted using Gabor-wavelet filters. Afterwards, the dimension of the features is reduced using principal component analysis (PCA). Finally a multi layer perceptron neural network is used to classify the facial expressions. The simulated results show high recognition rate as well as the low computational complexity that makes the proposed algorithm remarkable for accurate and fast facial expression recognition.

Original languageEnglish (US)
Title of host publicationHCI International 2013 - Posters' Extended Abstracts - International Conference, HCI International 2013, Proceedings
PublisherSpringer Verlag
Pages362-366
Number of pages5
EditionPART I
ISBN (Print)9783642394720
DOIs
StatePublished - 2013
Event15th International Conference on Human-Computer Interaction, HCI International 2013 - Las Vegas, NV, United States
Duration: Jul 21 2013Jul 26 2013

Publication series

NameCommunications in Computer and Information Science
NumberPART I
Volume373
ISSN (Print)1865-0929

Conference

Conference15th International Conference on Human-Computer Interaction, HCI International 2013
Country/TerritoryUnited States
CityLas Vegas, NV
Period7/21/137/26/13

Keywords

  • Facial expression recognition (FER)
  • Gabor-wavelet filters
  • Multi layer perceptron neural network
  • Wavelet-based salient point

ASJC Scopus subject areas

  • Computer Science(all)
  • Mathematics(all)

Fingerprint

Dive into the research topics of 'Automatic facial expression recognition using modified wavelet-based salient points and gabor-wavelet filters'. Together they form a unique fingerprint.

Cite this