Expression Recognition Methods Based on Feature Fusion

Expression recognition is popular research focus in Artificial Intelligence and Pattern Recognition. Feature fusion is one of the most important technical methods in expression recognition. To study how the feature information extracted from different part of the face play the role in facial expression recognition, experiments have been done and shown that Gabor wavelet feature and geometric characteristics of mouth are more important. In the first experiment, Gabor wavelet features of mouth is used for expression recognition, it is only worse than the result of the whole face. It has even better performance in Occidental emotion expression recognition. In the second experiment, we show that fusing the Gabor wavelet feature and geometric characteristics of mouth together can achieve better recognition results than using either method alone. It also has better real-time performance than using the whole face image.