Facial Expressions with Some Mixed Expressions Recognition Using Neural Networks

摘要

Facial feature extraction is the essential step of facial expression recognition. The automatic facial impression evaluation applies for wide area use. The important facial feature vectors for expression analysis are analyzed. The extracted feature vector loads all known feature vectors and trains the NN using as input training vectors while PCA is used for dimensionality reduction. The method is effective for both dimension reduction and good recognition performance in comparison with other proposed methods as shown in experiment results.