Using CNNs for American Sign Language fingerspelling recognition.
This Project was developed using python 3.x
We acquired our training from a public dataset published by the University of Massey. The dataset consists of 2524 close-up images of ASL fingerspelling with black backgrounds.
You will need python 3.x, and the following packages
Numpy
OpenCV
TensorFlow
Once you have downloaded the dataset, extract all the images into one folder (called images), in the same directory as the code in src. Then run from the same directory:
python alex_net2.py
Also, make sure to review the documentation to learn more about the approach taken in this project.