Fuzzy Wavelet Network with Feature Fusion and LM Algorithm for Facial Emotion Recognition

Abstract

With the development of face research, facial emotion recognition has become more and more important, and it has become the research hotspot now. In this paper, a fuzzy wavelet neural network (FWN) structure is proposed, in which the wavelet neural network is embedded into the fuzzy neural network. In the stage of feature extraction, features of face images in time domain and frequency domain are integrated to get more perfect information. In the stage of training and identification, we adopt the fuzzy wavelet network for classification, and levenberg-marquardt algorithm is utilized to accelerate the training speed. The experiment is tested on a facial emotion database to identify six emotions: neutral, sadness, anger, joy, disgust, surprise, fear. The experimental results show that the model can achieve better results.