- for running the program on the GPU (CUDA support) you need to build opencv from source with cmake (https://www.youtube.com/watch?v=tjXkW0-4gME)
- otherwise just
pip install -r requirements.txt
- run
getModels.sh
to download the predefined models - after all requirements are installed, there are many programs that can be run
take_a_video.py
this allowes the user to make a video to be processed laterbiceps_main.py
in which you can specify which video to process for the biceps exercisesquat_main.py
in which you can specify which video to process for the squat exerciseCamera2.py
where you have steps 1. and 3. combined, so, it takes the video, after you finish, its processes the video and gives you the output for the squat exercise
- all videos will be saved in the
videos
folder
Wrong squat where knees were too close.
Correct squat.
Biceps example.