基于深度学习的面部表情识别研究

针对基于深度学习的静态人脸图像表情识别方法进行研究，首先介绍了深度学习的原理，并归纳了目前公开且常用的面部表情数据集；然后介绍了基于深度学习的表情识别的三个步骤，归纳了图像预处理和表情分类的主要方法，重点总结了目前性能较好用来提取特征的深度学习框架以及这些方法的基本原理和优劣势比较；最后指出了目前面部表情识别存在的问题和未来可能的发展趋势。

Static face image expression recognition method based on deep learning

This paper studied the static face image expression recognition method based on deep learning. Firstly, it introduced the principle of deep learning, and summarized the current public and commonly used facial expression data sets. Then it introduced the three steps of expression recognition based on deep learning. It summarized the main methods of image preprocessing and expression classification. This paper mainly summarized the deep learning frameworks with better performance to extract features, the basic principles, advantages and disadvantages of these methods. Finally, the paper pointed out the problems of facial expression recognition and possible future development trends.