Inspiration

We wanted to be able to learn languages with unreliable tools slowing us down. So we invented a companion anyone could use to help them learn a new language.

What it does

The LinguaPet sits on your desktop and waits for the user to ask it something. It then creates a response in the users native language and their target language. Both of these are displayed on the arduino so the user can improve their reading of their target language. Alongside this the program reads out the target language message to the users so they can improve their listening ability. It also has some built in language games that test the users ability to actually speak their target language.

How we built it

On the software side the program puts itself in a cycle where it waits for the user to prompt it for a request. It then listents to the mic until the user signals that thier reqeust is over. Once the request has been processed from the mic it uses Eleven labs to parse that audio into text. Then it passes the text into google gemini so it can create a response in the users native language and their target language. Once this response has been created it is sent over the network to the arduino so it can display both languages on the display and sends the target language text to Eleven labs text to speech and playback so the user can practice listening to their target language alongside reading it.

Challenges we ran into

Getting the hardware to work. Getting the software to work. Getting anything to work.

Accomplishments that we're proud of

The fact that the hardware works. The fact that our pipeline for receiving, parsing, translating, and presenting the final message to the user all work. The fact that we managed to get it done.

What we learned

How to use eleven labs, how to use rasberry pi's and LCD screens, and virtual python environments.

What's next for LinguaPet

A cute pet body to surround the components instead of them being exposed, a better lcd display so we can display things beyond ascii, the ability to learn the users tendencies and inflection over time.

Built With

Share this project:

Updates

Submission history