Facial Expression Recognition Based on Selective Feature Extraction

Abstract

As one of the key techniques for futuristic human computer interaction, facial expression recognition has received much attention in recent years. A method of facial expression recognition based on selective feature extraction is presented in this paper. In this method we classify expressions roughly into three kinds according to the deformation of mouth firstly. Then we select some expression feature areas which contribute much to each kind expression according to the rough classification results and extract features for them. Lastly we classify expressions finely using method based on rule. Experiments show that facial expression recognition based on selective feature extraction can get high recognition rate and has strong robustness