Real-time Sign Language Translator

A prototype application that uses a webcam to capture hand gestures in real-time and translates them into text. The system is built with Python, Flask, and OpenCV, and it leverages a pre-trained TensorFlow/Keras model to recognize and interpret Indian Sign Language (ISL) gestures.

• Used Flask to create a lightweight web server for real-time video streaming.
• Leveraged OpenCV for efficient webcam access and real-time image processing.
• Integrated a pre-trained TensorFlow/Keras model for sign language gesture recognition.
• Designed a simple HTML frontend to display the video feed and classification results without the overhead of a complex JavaScript framework.

1. Set up the Flask application to handle video streaming.
2. Integrated OpenCV to capture and process webcam frames.
3. Loaded the pre-trained Keras model (.h5) for predictions.
4. Developed the logic to preprocess the ROI and feed it to the model.
5. Designed the simple HTML interface to display the results.