We wanted to challenge ourselves with technologies we have never used. We thought that it would be beneficial to everyone to be able to video call and speak in the language of their choice whether it is for business or personal reasons. Canada Heritage's challenge on solving issues that many anglophone speakers face in Quebec also inspired us to tackle this issue.
What it does
It establishes a peer-to-peer video call that automatically translates in real-time the languages being spoken. It also accepts sign languages to the webcam as input.
How we built it
Using Google Cloud for translation and WebRTC for establishing a video call. The web app is built using Ionic and can therefore be deployed to mobile devices. The sign language recognition is being handled by a custom machine learning server.
Challenges we ran into
Streaming with WebRTC, machine learning training.
Accomplishments that we're proud of
Have a working product that we can proudly show to everyone.
What we learned
Machine learning, WebRTC connections.
What's next for Tchat
Improving the electronic voice and improving the machine learning dataset.