An LSTM model that recognizes alphabets from the American Sign Language using Mediapipe and Tensorflow. It uses mediapipe to get landmarks of the left and right hands and passes it through the LSTM model.
- Tensorflow
- Mediapipe
- OpenCV
- NumPy
- scikit-learn
Installation: pip install tensorflow opencv-python scikit-learn numpy mediapipe
After running the python notebook we get a webcam stream with the predicted alphabet displayed in the window as below:
