ASLingo

• 

  Main homepage for ASLingo.

• 

  Learning at it's finest.

Inspiration

ASLingo, a platform designed to dismantle communication barriers through the power of American Sign Language (ASL). Our idea was simple yet challenging: to make ASL accessible, engaging, and enjoyable for everyone.

What it does

ASLingo broadcasts a webcam of the user and quizzes them on various signs by sending webcam data to the Flask/TensorFlow backend

How we built it

Our techstack primarily consists of TensorFlow, Flask, Vanilla JS, HTML/CSS. To be more specific, we started at some documentation, and after a caffeine-fueled night, the code seemed to magically write.

Challenges we ran into

Web servers are tough, especially when using a new framework like Flask. TensorFlow was hard to get a grasp on, which led to a hard crunch overnight and getting it to interact with the rest of our web app.

Accomplishments that we're proud of

Learning TensorFlow and Flask all in 1 night, and this being the first project we made in Python this was already a challenge we put up to ourselves. Training LLMs and getting it to communicate a JavaScript frontend was particularly challenging.

What we learned

How to create a web server using Flask, how to train LLMs in TensorFlow, how to read webcam data from a JS, and also managing dependencies with VENV in Python. Needless to say, we learned a lot, and it all came from purposely choosing a techstack we weren't at all familiar with.

What's next for ASLingo

We want to regularly get the snapshots to be processed in under 10 seconds regularly, and also train the model on a more diverse set of people.

Built With

• flask
• htmlcss
• javascript
• numpy
• python
• tensorflow