基于MobileNetV2与LBP特征融合的婴幼儿表情识别算法  

Infant Expression Recognition Algorithm Based on MobileNetV2 and LBP Feature Fusion

作　　者：邓源 施一萍[1] 江悦莹 朱亚梅 刘瑾[1] DENG Yuan;SHI Yiping;JIANG Yueying;ZHU Yamei;LIU Jin(School of Electronic and Electrical Engineering,Shanghai University of Engineering Science,Shanghai 201620,China)

机构地区：[1]上海工程技术大学电子电气工程学院,上海201620

出　　处：《电子科技》2022年第8期47-52,共6页Electronic Science and Technology

基　　金：国家自然科学基金(61701296);上海工程技术大学研究生科创项目(20KY0218)。

摘　　要：针对婴幼儿表情识别率低、特征复杂提取不充分等问题,文中提出一种基于MobileNetV2与LBP双通道特征融合的婴幼儿表情识别算法。第1条通道使用改进后的MobileNetV2网络,可快速、准确地提取出人脸表情全局特征。第2条通道对原始输入图进行分块,利用图像信息熵构造出权值,提取出分块加权LBP直方图特征,突出了表情信息丰富的区域。通过融合双通道模型的输出向量来提升特征表达能力,并采用支持向量机替代Softmax层进行表情分类。实验表明,使用融合特征比单一特征具有更好的分类效果,并且在自建的婴幼儿表情数据集中的表情识别准确率可达到85.71%。In view of the problems of low rate of infant expression recognition and the insufficient extraction of complex features,an infant expression recognition algorithm based on MobileNetV2 and LBP dual-channel feature fusion is proposed.The first channel uses the improved MobileNetV2 network to quickly and accurately extract the global features of facial expressions.The second channel divides the original input image into blocks,and uses image information entropy to construct weights,and extracts block-weighted LBP histogram features to highlight the regions with rich expression information.The output vector of the dual-channel model is fused to improve the feature expression ability,and the support vector machine is used to replace the Softmax layer for expression classification.Experiments show that the use of fusion features has a better classification effect than a single feature,and in the self-built infant expression data set,the accuracy of expression recognition can reach 85.71%.

关 键 词：卷积神经网络 局部二值模式 特征融合 双通道模型 表情识别 图像信息熵 婴幼儿 支持向量机 

分 类 号：TP391[自动化与计算机技术—计算机应用技术] TN99[自动化与计算机技术—计算机科学与技术]