A New Recursive Neural Network and Center Loss for Expression Recognition

Dongling Wang ； Xiaowei Wang ； Shoulin Yin

《International Journal of Electronics and Information Engineering》 13卷3期 (2021/09) Pp. 97-104

https://doi.org/10.6636/IJEIE.202109_13(3).02

摘要

A novel facial expression recognition algorithm with a recursive neural network and central loss function based on a convolutional neural network is proposed to solve the problem of low recognition accuracy in facial expression recognition. The recursive neural network is used to learn the features of different feature channels to improve the model's sensitivity to the feature channels and further enhance the feature representation ability of the model. At the same time, the central loss function combined with the Softmax loss function is used to supervise and train the neural network, effectively increasing the distance between classes, ensuring the in-class tightness, and improving the recognition accuracy of the model. By testing on the CK+ dataset, the experimental results show that the recognition accuracy of the improved model reaches 98.12%, which is higher than other typical algorithms.

關鍵字

Central Loss Function ； Facial Expression Recognition ； Recursive Neural Network