The code can be used to construct a Detector object which will detect the hand sign for the letter Y through a camera
Multiple neural networks are connected together in order to analyze images
If the hand sign for Y is detected, then final_application.py opens youtube.com
Run training_data_collector.py and test_data_collector.py to collect training and test data
After that run subnet_trainer.py to generate a Network pickle
If the accuracy is high enough, save the pickle by giving it a unique name and moving it to the network_pickles folder.
Run ensemble_training_data_collector.py with the proper initialization of the variables.
Run intermediate_network_trainer.py.
Rename and save the Network pickle.
In y_sign_detection.py make sure the Detector is instantiated with the appropriate sub and intermediate Network objects.
Run the script and it should return the final accuracy of the Detector.
Below the accuracy, there should also be an message that says "Detector Saved"
If you get a MemoryError, try commenting out the part of the script that evaluates the Detector and run again.
Once you have a final detector pickle, run final_application.py.
Whenever you present the Y hand sign at the camera, youtube.com will open up.
Part 1: https://youtu.be/GdPtevjUfzo
Part 2: https://youtu.be/nhyEixTzuDg
Part 3: Coming soon ...