The ability to easily communicate with others is something that most of take for granted in our everyday life. However, for the millions of hearing impaired and deaf people all around the world, communicating their wants and needs is a battle they have to go through every day. The desire to make the world a more accessible place by bringing ASL to the general public in a fun and engaging manner was the motivation behind our app.

What it does

Our app is essentially an education platform for ASL that is designed to also be fun and engaging. We provide lessons for basic ASL such as the alphabet, with plans to introduce more lessons in the future. What differentiates our app and makes it engaging, is that users can practice their ASL skills right in the app, with any new letter or word they learn, the app uses their webcam along with AI to instantly tell users when they are making the correct sign. The app also has a skills game that puts what they learnt to the test, in a time trial, that allows users to earn score for every signed letter/word. There is also a leaderboard so that users can compete globally and with friends.

How we built it

Our app is a React app that we built with different libraries such as MUI, React Icons, Router, React-Webcam, and most importantly Fingerpose along with TensorflowJS for all our AI capabilities to recognize sign language gestures in the browser.

Challenges we ran into

Our main struggle within this app was implementing Tensorflowjs as none of us have experience with this library prior to this event. Recognizing gestures in the browser in real time initially came with a lot of lag that led to a bad user experience, and so it took a lot of configuring and debugging in order to get a much more seamless experience.

Accomplishments that we're proud of

As a team we were initially building another application with a similar theme that involved hardware components and we had to pivot quite late due to some unforseen complications, and so we're proud of being able to turn around with such a short amount of time and make a good product that we would be proud to show anyone. We're also proud of building a project that also has a real world usage to it that we all feel strongly about and that we think really does require a solution for.

What we learned

Through this experience we all learned more about React as a framework, in addition to real time AI with Tensorflowjs.

What's next for Battle Sign Language

Battle Sign Language has many more features that we would look to provide in the future, we currently have limited lessons, and our gestures are limited to the alphabet, so in the future we would increase our app to include more complex ASL such as words or sentences. We would also look forward to adding multiplayer games so that people can have fun learning and competing with friends simultaneously.

Built With

Share this project: