基于卷积神经网络的学生课堂面部表情识别研究

Study on Facial Expression Recognition in Class Based on Convolutional Neural Network

学生课堂面部表情识别产品不适合采用“大样本”的研发方法,为减少数据采集工作量,提出使用卷积神经网络算法,分析了卷积神经网络基本原理,探讨了总体结构,设计了识别流程,以证件照为原始样本构建了人脸数据集,以表情特征训练的方法扩展数据集,经实验研究表明,该方法表现良好,能够有效开展识别。

Students'facial expression recognition products are not suitable for the"large sample"research and development method,in order to reduce the workload of data collection,convolutional neural network algorithm is proposed,the basic principle of convolutional neural network is analyzed,the overall structure is discussed,the recognition process is designed,the face data set with the original sample of certificate photo is constructed,and the data set is extended to the method of expression feature training.The experimental results show that this method performs well and can workeffectively in recognition.