I submitted this project as Swift Student Challenge for WWDC'23.
The main objective of this application is to help parents learn ASL for their children who are born deaf or not able to speak. You can also convert your actions to ASL Letters Actions to English Alphabets.
Feel free to explore the project and give feedback.
- You need to install Swift Playgrounds for iPad/MacOS.
- You need to give camera permission.
The machine learning model is not the perfect. I trained the Hand Action Classifier Model with the CreateML
The code is also not the perfect since i coded the whole application with the model training in 2 weeks. I didn't consider the good practice because of short deadline also considering the complexity of idea.
https://www.kaggle.com/competitions/asl-signs/data
https://www.kaggle.com/competitions/asl-signs/code
https://www.kaggle.com/datasets/risangbaskoro/wlasl-processed
https://www.kaggle.com/datasets/debashishsau/aslamerican-sign-language-aplhabet-dataset