2020 Eng Fair Project. Justin Moonjeli and Nevin Gregory. Won 1st place in Gwinnett County Regional Fair Computer Systems.
The purpose of this project was to provide a proof-of-concept that showed a way for easier communication for those with verbal disabilities. This was to be accomplished in the form of an app or program that could detect and translate sign language spelled into a camera and output it as text. In order to accomplish this, a dataset consisting of the American Sign Language (ASL) alphabet was imported. Then a neural network was written which would use the 28x28 images from the dataset, analyze it, and output a letter. The network was trained and then tested on new images in which it performed successfully. The data came from the training stage and showed the improvement of accuracy as the repetitions increased. The neural network was able to reach a final accuracy of approximately 99%. In conclusion, the prototype that was built is a starting point that shows that the idea is viable and can be used to further ease of communication