Optical-Flow Based Symmetric Feature Extraction for Facial Expression Recognition

EasyChair Preprint 10394

Abstract

Facial expression analysis is one of the most important tools for behavior interpretation and emotion modeling in Intelligent Human-Computer Interaction (HCI). Although humans can easily interpret facial emotions, computers have great difficulty doing so. Analyzing changes and deformations in the face is one of the methods through which machines can interpret facial expressions. However, maintaining great precision while being accurate, stable, and quick is still a challenge in this field. To address this issue this research presents an innovative and novel method to extract key features from a face during a facial expression fully automatically. These features can be used by various machine learning models to analyze emotions. We used the optical flow algorithm to extract motion vectors, which were then divided into sections on the subject’s face. Finally, each section and its symmetric section were used to calculate a new vector. The final features produce a state-of-the-art accuracy of over 98% in emotion classification in the Extended Cohen-Kanade (CK+) facial expression dataset. Furthermore, we proposed an algorithm to filter the most important features, and with an SVM classifier, we were able to keep the accuracy over 98 % by only looking at 10% of the face area.

Keyphrases: Extended Cohen-Kanade (CK+) Dataset, Optical Flow Algorithm, emotion recognition, facial expression recognition, feature extraction

@booklet{EasyChair:10394,
  author    = {Mohammad Ali Zeraatkar and Javad Hassannataj Joloudari and Kandala N V P S Rajesh and Silvia Gaftandzhieva and Sadiq Hussain},
  title     = {Optical-Flow Based Symmetric Feature Extraction for Facial Expression Recognition},
  howpublished = {EasyChair Preprint 10394},
  year      = {EasyChair, 2023}}