参考文章:
吸烟行为检测:
yolov3+tensorflow+keras实现吸烟的训练全流程及识别检测:
https://blog.csdn.net/liuwuw/article/details/103630996
https://github.com/swliu2016/train-core-yolov3-darknet
darknet实现吸烟的训练全流程及识别检测:
https://blog.csdn.net/liuwuw/article/details/104188246
脸部疲劳驾驶检测:
基于Python实现的疲劳驾驶检测:
https://www.write-bug.com/article/2509.html
疲劳检测(Fatigue Detection Algorithm):
https://blog.csdn.net/laiwenjie1988/article/details/85276151
基于机器视觉的疲劳驾驶检测:
https://blog.csdn.net/qq_31065851/article/details/92012235
Dlib模型之驾驶员疲劳检测二(打哈欠)(有完整代码,但判断方法较差)
https://blog.csdn.net/cungudafa/article/details/103496881
头部姿态估计:(有OpenPose、OpenFace、AlphaPose、MaskRCNN解决方案)
OpenPose源码:
https://github.com/CMU-Perceptual-Computing-Lab/openpose
https://github.com/Hzzone/pytorch-openpose
OpenFace源码:
https://github.com/TadasBaltrusaitis/OpenFace
https://github.com/thnkim/OpenFacePytorch
AlphaPose源码:
https://github.com/MVIG-SJTU/AlphaPose
https://github.com/Amanbhandula/AlphaPose
MaskRCNN源码:
https://github.com/facebookresearch/detectron2
重磅!头部姿态估计「原理详解 + 实战代码」来啦!
https://www.sohu.com/a/278664242_100007727
OpenPose:实时多人2D姿态估计(详细,还有定义)
https://blog.csdn.net/qq_41204464/article/details/90656950
AlphaPose论文笔记《RMPE: Regional Multi-person Pose Estimation》
https://www.cnblogs.com/taoshiqian/p/9593901.html
人体姿态估计Alphapose配置安装教程(GPU,超详细,亲测有效!)
https://blog.csdn.net/qq_38522972/article/details/82993216
Openface人脸朝向和注意力检测
https://blog.csdn.net/where_is_horse/article/details/80099124
计算机视觉中头部姿态估计的研究综述(概念较多,但写得过于简单)
https://blog.csdn.net/weixin_41703033/article/details/83215043
Landmark估计头部姿态
使用opencv和dlib进行人脸姿态估计(python)(欧拉角计算有问题,飘的厉害,且检测到的面部较高,无工业化价值)
https://blog.csdn.net/yuanlulu/article/details/82763170
head-pose-estimation-using-opencv-and-dlib(上一篇的原文)
https://www.learnopencv.com/head-pose-estimation-using-opencv-and-dlib/
基于OpenCV和Dlib的头部姿态估计[译]
https://www.aiuai.cn/aifarm909.html
基于Dlib和OpenCV的人脸姿态估计(HeadPoseEstimation)(计算到欧拉角的办法)
https://blog.csdn.net/u013512448/article/details/77804161
人脸姿态估计浅谈~(方法较老,不实践)
https://blog.csdn.net/wsj998689aa/article/details/39294171
欧拉角、旋转向量和旋转矩阵的相互转换
https://www.jianshu.com/p/5e130c04a602
300W-LP、AFLW2000、BIWI数据集:
AFLW2000-3D数据库介绍及自带代码使用
https://blog.csdn.net/AuntieLee/article/details/105940291
http://www.cbsr.ia.ac.cn/users/xiangyuzhu/projects/3DDFA/main.htm
https://www.kaggle.com/kmader/biwi-kinect-head-pose-database/data
aflw数据集使用
https://blog.csdn.net/weixin_30558305/article/details/97031117
驾驶员行为识别:
基于深度学习的3D骨架动作识别
https://blog.csdn.net/u013608336/article/details/81070645
基于深度学习的Action Recognition(行为识别)【一】
https://blog.csdn.net/whfshuaisi/article/details/79116246
基于深度学习的Action Recognition(行为识别)【二】
https://blog.csdn.net/whfshuaisi/article/details/79116265
[骨架动作识别]A New Representation of Skeleton Sequences for 3D Action Recognition
https://blog.csdn.net/u013608336/article/details/78740343
驾驶员活体检测:
人脸识别中的活体检测算法综述
https://sigai.blog.csdn.net/article/details/82772813
相关数据集:
CEW眼睛状态数据集
http://parnec.nuaa.edu.cn/xtan/data/ClosedEyeDatabases.html
打哈欠数据集
http://www.site.uottawa.ca/~shervin/yawning/
Kaggle驾驶员分心检测竞赛:
https://www.kaggle.com/c/state-farm-distracted-driver-detection/data
代码实现:
https://www.kaggle.com/ismailchaida/cnn-to-detect-driver-actions
https://github.com/omnigeeker/mlnd_distracted_driver_detection
github上有很多相关工程