Robust facial expression recognition in the presence of rotation and partial occlusion

Authors:
Diego Mushfieldt;Mehrdad Ghaziasgar;James Connan
Affiliations:
University of the Western Cape, Bellville, Cape Town;University of the Western Cape, Bellville, Cape Town;Rhodes University, Grahamstown
Venue:
Proceedings of the South African Institute for Computer Scientists and Information Technologists Conference
Year:
2013

Citing 6
Cited 0

Multiresolution Gray-Scale and Rotation Invariant Texture Classification with Local Binary Patterns

IEEE Transactions on Pattern Analysis and Machine Intelligence
Robust Real-Time Face Detection

International Journal of Computer Vision
Local versus Global Segmentation for Facial Expression Recognition

FGR '06 Proceedings of the 7th International Conference on Automatic Face and Gesture Recognition
An analysis of facial expression recognition under partial facial image occlusion

Image and Vision Computing
Eye Detection Algorithm on Facial Color Images

AMS '08 Proceedings of the 2008 Second Asia International Conference on Modelling & Simulation (AMS)
Local binary patterns for multi-view facial expression recognition

Computer Vision and Image Understanding

Quantified Score

Hi-index	0.00

Visualization

Abstract

The research presented in this paper proposes an approach to recognizing facial expressions in the presence of rotations and partial occlusions of the face. The research is in the context of automatic machine translation of South African Sign Language (SASL) to English. The proposed method achieved a high average recognition accuracy of 85% for frontal facial images. It also achieved a high average recognition accuracy of 80% for faces rotated at 60°. It was also shown that the method is able to continue to recognize facial expressions even in the presence of full occlusions of the eyes, mouth and left and right sides of the face. An additional finding was that both the left and the right sides of the face are required for recognition. This was due to the fact that subjects are seen to regularly perform facial expressions with more emphasis on either side, affecting the recognition accuracy.