基于深度学习的危险驾驶检测算法python源码+视频（判断是否闭眼或者张开嘴哈欠和吸烟打电话等手势行为等）.zip

本项目是用于判断是否闭眼或者张开嘴哈欠和吸烟打电话等手势行为， 功能涵盖7类情绪识别，眨眼判断，哈欠判断，吸烟，打电话等， 达到危险驾驶检测的功能。 项目参考https://blog.csdn.net/Aristochi/article/details/105655810 ``` 数据集：闭眼、睁眼、闭嘴、哈欠、打电话、吸烟数据集 口罩数据集 fer2013表情数据集 Data set: closed eyes, open eyes, yawn, telephone, smoking data set Mask data set Fer2013 Emoji Data Set ``` 眨眼数据集http://parnec.nuaa.edu.cn/xtan/data/ClosedEyeDatabases.html 数据集 链接：https://pan.baidu.com/s/1FLfWZix-vsY5XSaV1oi-pA 提取码：pmh0 ``` Done: 眨眼判断 哈欠判断 打电话识别 吸烟识别 预测准确率：93.46% Done: Blink of an eye to judge Yawn judgment Telephone identification Smoking is to identify Prediction accuracy: 93.46% ``` TODO： PERCLOS疲劳指标判断 将项目移植到树莓派 TODO: PERCLOS fatigue index judgment Port the project to raspberry pie Thanks for https://github.com/ipazc/mtcnn/tree/master/mtcnn MTCNN module 感谢https://github.com/ipazc/mtcnn/tree/master/mtcnn的MTCNN模块 ```