Automatic Facial Expression Recognition Using Fusion of Feature Extraction Techniques

Authors: Arpita Waze, Deepak Sharma

Certificate: View Certificate

Abstract

With the increasing immersion of computers in our everyday life, the gap between computers and humans becomes increasingly apparent. Face plays significant role in social communication. This is a \'window\' to human personality, emotions and thoughts. Facial expressions, resulting from movements of the facial muscles, are the face changes in response to a person’s internal emotional states, intentions, or social communications. The three stages of facial expression recognition are pre-processing, features extraction and classification. Feature extraction plays an important role in facial expression recognition. To enhance the recognition accuracy of the facial expression recognition we adopt the merging approach which combines different feature extraction technique for local feature extraction improvement. In this paper, we are evaluating 2DPCA+LBP for facial representation. To improve the accuracy of the system we have applied 2DPCA on LBP images in place of original images. The proposed system has achieved the recognition rate of 97.25 % for 2DPCA+LBP.

Introduction

Facial expressions, resulting from movements of the facial muscles, are the face changes in response to a person’s internal emotional states, intentions, or social communications. there's a substantial history related to the study on facial expressions. Darwin (1872) was the primary to explain in details the precise facial expressions related to emotions in animals and humans, UN agency argued that every one mammals show emotions dependably in their faces. Since that, facial features analysis has been a space of nice analysis interest for activity scientists (Ekman, Friesen, and Hager, 2002). Psychological studies (Mehrabian, 1968; Ambady and Rosenthal, 1992) recommend that facial expressions, because the main mode for non-verbal communication, play a significant role in human communication.Facial expression recognition needs a lot of delicate and discriminative feature extraction as compare to alternative recognition ways.

Conclusion

The complete study concludes that; to take care of the equilibrium in choosing helpful info and reducing unwanted info or lower face regions, we\'ve got any applied Adaboost methodology to urge the foremost necessary info from a face image i.e. the central region of the face composed of Eyes, Nose and Mouth. To derive the importance of facial elements we\'ve got used the foremost necessary elements of face as module of 2DPCA. it\'s not solely four elements of face image however these are four identification pillars that are Left and Right Eye, Nose and Mouth. Moreover, few expressions appear to be a tough to properly classify. This principally results from the actual fact that the performance of expressions varies among subjects. The experiments demonstrate that that facial half plays necessary role in classification of explicit expression. as an example, angry expression are often recognized properly with facilitate of 2 facial elements i.e. Eyes and Mouth .In same manner we tend to known the facial elements that have significance in recognition of expressions. the whole work has been simulated and evaluated by MATLAB.