Detection of Facial Landmarks from Neutral, Happy, and Disgust Facial Images

Abstract

Automated analysis of faces showing different expressions has been recently studied to improve the quality of human-computer interaction. In this framework, the expression-invariant face segmentation is a crucial step for any vision-based interaction scheme. A method for detecting facial landmarks from neutral and expressive facial images was proposed. In present study, a particular emphasis was given to handling expressions of happiness and disgust. The impact of these expressions on the developed method was tested using dataset including neutral, happiness and disgust images. The results demonstrated a high accuracy in detecting landmarks from neutral images. However, the expressions of happiness and disgust had a deteriorating effect on the landmark detection.