基于深度学习的人脸表情分析

广东东软学院本科生毕业设计（论文） 基于深度学习的人脸表情分析 内容摘要人脸表情是人类在没有语言交流时的一种表达感情的方式，而且是最常见与最容易表达的，对于计算机来说，所谓的人脸表情识别就是输入一张图片给计算机，计算机通过一些特征提取，数值计算，最后得出表情的类别，面部表情识别是基于人工智能的人机交互技术的重要组成部分。表情识别在现实中应用的场景甚多，例如零售业中通过识别顾客的表情，获取他对商品的喜好，或者在游戏中，可以识别用户的表情，进行人机交互体验等等。本文将会利用目标检测算法 YOLOV3 和图像识别算法 VGGNET 对图片中的人进行一个表情识别。关键词： 深度学习 神经网络 特征提取 表情识别 目标检测 目标识别 卷积神经网络广东东软学院本科生毕业设计（论文） Abstract The recognition of facial expressions is that artificial intelligence is an important part of human-computer interaction technology. Facial expressions are one of the richest resources in our daily non-verbal communication and the easiest way. Expressions are expressed in a certain way. For a computer, facial expression recognition is an image input by a computer. The computer has some feature extraction, numerical calculation, and category analysis expressions.There are many scenarios in which facial expression recognition is applied in reality. For example, in the retail industry, customers' facial expressions can be recognized to obtain their preferences for goods, or in the game, users' facial expressions can be recognized to conduct man-machine interaction experience.In this paper, target detection algorithm YOLOV3 and image recognition algorithm VGGNET will be used to carry out an expression recognition of people in the picture.Key words： Deep learning Neural Networks Feature extraction Expression recognition Target Detection 广东东软学院本科生毕业设计（论文）Target Recognition Convolutional Neural Network 目录一、 绪论.........................