To keep connected people, using the resources that everyone has
What it does
A sign language translator that uses a webcam, and translates some signs to be more accessible to everyone
How we built it
Using python and python libraries (as OpenCV, Tensorflow, Image, etc) and azure technologies.
Challenges we ran into
How to identify one frame of the video, track the sign and do the approximation
Accomplishments that we're proud of
It runs without internet, and from a local way
What we learned
How to implement Tensorflow to the edge (to run offline and in a local way) and to implement Tensorflow with a video stream
What's next for Hands-On!
Try to upgrade the Azure model, to be either more precise and to have more signs