Trained a neural network to recognize American Sign Language (ASL) gestures using Keras and OpenCV.
Built and trained a convolutional neural network for the recognition of American Sign Language (ASL) gestures. The network was trained on a Kaggle dataset of ASL gestures uploaded by Ayush Thakur.
The model attained a 57% accuracy on a test set extracted from the dataset.
To run the code install the requirements with python 3 by
pip install -r requirements.txt
After this, the tool can be run with python main.py
The results of the engine are displayed in the application window.