Script that employs yolov8s-pose to detect humans in an image, retrieve the width and height of their bounding boxes, and ultimately determine if the width/height ratio exceeds 1.4, indicating a potential fall. It has a %96.3 accuracy rate,and runs pretty fast with CUDA. (For Nvidia RTX 3050 Mobile:385 FPS) Dataset link: https://falldataset.com/ (i know this is not that user friendly but i just wanted to put this here in case it helps someone)
-
Notifications
You must be signed in to change notification settings - Fork 2
andmydignity/fall_detection_yolov8s
Folders and files
Name | Name | Last commit message | Last commit date | |
---|---|---|---|---|
Repository files navigation
About
No description or website provided.
Topics
Resources
Stars
Watchers
Forks
Releases
No releases published
Packages 0
No packages published