ASL Translate is a real-time American Sign Language (ASL) gesture translator. It uses a webcam to capture ASL gestures, processes them through a pre-trained model, and outputs the corresponding letter as text.

• Real-time ASL gesture recognition
• Web-based interface using JavaScript, CSS, and HTML
• Flask backend with a pre-trained image classification model
• Serves static React frontend files

• Frontend: JavaScript, CSS, HTML (React)
• Backend: Flask (Python)
• Machine Learning: Vision Transformer Model from Hugging Face Transformers, Image Classification Model (Marxulia/asl_aplhabet_img_classifier_v3)

Check out the hosted website: ASL Translate

If you'd like to run the application locally, follow these steps:

1. Clone the repository:

   git clone https://github.com/Yahialqur/ASL_Translate.git
   cd ASL_Translate
   pip install -r requirements.txt
   python app.py

   The frontend is served through flask, you can access it by visiting http://127.0.0.1:5000/ after running the backend.

• Ensure your webcam is connected and allowed in your browser
• Perform ASL alphabet gestures with your hand in front of the webcam
• The application will display the recognized letter in real time
• Keep in mind this is still a prototype so accuracy is not yet perfect