ASL Net API. A machine learning powered API that inteprets American Sign Language by Images. Built on a flask web server and model pretrained using Keras. Built by D. Romeno Wenogk Fernando, Takumi Miyawaki and Munachiso Nwadike. Demo available.
We made a custome dataset of 500 ASL Alphabet and trained a YOLOV3 model to be able to detect sign language using our integrated mobile application.
A detailed summary on our work can be found in the attached white-paper (.pdf) in this repository.