跌倒检测算法笔记

基于传感器的数据：rnn

https://github.com/chizhanyuefeng/Realtime-Fall-Detection-for-RNN

ssd

c4d跌倒检测ssd改进效果测试.rar

c4d跌倒检测ssd改进效果测试.rar_跌倒检测深度学习-深度学习工具类资源-CSDN下载

这俩好像一样

c4d跌倒检测ssd改进CL加速效果测试.rar_yolov5跌倒检测-深度学习工具类资源-CSDN下载

跌倒检测-OPENCV-VC++

这里面有跌倒视频

跌倒检测-OPENCV-VC++_opencv跌倒检测,跌倒检测-C++工具类资源-CSDN下载

Yolov3和光流估计版本：

https://github.com/jacke121/Fall-Detection-By-YOLOV3-and-LiteFlowNet

本人也做过一版关于姿态估计的跌倒检测

openpost的姿态判断跌倒：

https://github.com/BlackFeatherQQ/openpose_fall_detect

博文介绍：

跌倒识别 摔倒识别 -lightweight_openpose_sinat_28371057的博客-CSDN博客

RNN的：开源了数据，

https://github.com/chizhanyuefeng/Realtime-Fall-Detection-for-RNN

https://github.com/AdrianNunez/Fall-Detection-with-CNNs-and-Optical-Flow

vgg16

Dataset	Sensitivity	Specificity	FAR	MDR	Accuracy
URFD	99.67% (-+0.67%)	98.57% (-+0.56%)	1.43% (-+0.56%)	0.33% (-+0.67%)	98.63% (-+0.52%)

https://github.com/GajuuzZ/Human-Falling-Detect-Tracks

• Tiny-YOLO oneclass - .pth, .cfg
• SPPE FastPose (AlphaPose) - resnet101, resnet50
• ST-GCN action recognition - tsstg