Before you jump into executing the code create a conda environment including the necessary dependencies by running the following command:
conda env create --file environment.yml
The application works by running the TeachingSlate.ipynb and the video feed can be changed by the changing the videofeed parameter to cv2.VideoCapture(videofeed). The videofeed is currently from an app called as droidCam (https://play.google.com/store/apps/details?id=com.dev47apps.droidcam&hl=en_IN&gl=US)
The yolo model used for hand detection is taken from https://github.com/cansik/yolo-hand-detection
The model weights for cross hands tiny is available in https://github.com/cansik/yolo-hand-detection/releases/download/pretrained/cross-hands-tiny-prn.weights Download the model weights and place it in the models folder
The setup to the yolo model is briefly explained in https://github.com/cansik/yolo-hand-detection/blob/master/README.md
Credits: Florian Bruggisser https://github.com/cansik