This repository is to be used as a learning aide for American Sign Language. This web application uses Angular for the front-end and Flask for the back-end
Run docker compose watch which will start a web application that consists of an Angular front-end (default port 4200) and a Flask back-end
In your browser navigate to http://localhost:4200 (or whatever port you defined for the front-end in docker-compose.yml)
In the flask-backend container, run python ./model.py to train the model. The model will save in the directory as asl_model.keras. Use docker cp to save the model to your host machine.
Training analysis is saved to the directory as analysis.png , use docker cp to save the .png file to your host machine to view training results.
Add asl_model.keras file to /backend directory
The dataset used for this project was not created by myself. It is available on Kaggle as the ASL Alphabet Dataset. https://www.kaggle.com/grassknoted/asl-alphabet
