Despite being the only effective form of communication for many people across the globe, sign language is often ignored or forced to take a back seat in light of other spoken languages. We believe this should change, that learning sign language should be a fun and engaging experience, so we developed a game to do just that.
What it does
FingerAlphabet is a fast-paced, augmented reality single-player game, designed to teach sign language. Once the player selects a segment of speech to play, they will be put against the clock to sign it out as quickly and accurately as possible, aiming to get the highest score.
How was it built
- Leap Motion - This detects the precise movements of the player's hands
- Microsoft Azure - We used a Decision Forest machine learning module from Azure to check if the detected position is correct
Challenges we ran into
- CORS on Azure was somewhat confusing
- Figuring out the optimum range of the Leap took some time.
Accomplishments that we're proud of
- Making a completed product, functional product!
iF wE HaD TImE
Given the time, we would develop the game further by:
- adding extra levels of difficulty, e.g. making it harder by not showing the player the sign for the word
- adding a multiplayer feature for more competitive gameplay
- allowing the player to supply their own text, which would require a much larger library of signs but make the game more customisable; related to this is also having the ability for the player to add more signs