Use collect_imgs.py to collect the data on which the model will be trained. Collect_imgs.py script opens front camera and takes photos of the user which should make the corresponding sign language character with his right hand (A, B and then L). Make sure to move your hand towards and from camera during the data collecting as to assure data diversity and better training.