EchoStage

*WE WILL EDIT THIS ONCE WE ACTUALLY GET A WORKING DEMO*

Inspiration

Actors traditionally need to perform in very different roles that often require different accents, but training for this is time-consuming and doesn't have easy feedback if there is not a coach. We saw a chance to expand these opportunities with AI.

What it does

This tool compares one's accent of a prompt with an AI-generated accent, and gives both a score of how well the accent compares and feedback on what words to improve upon.

How we built it

We used ElevenLabs for generated a text-to-speech AI accent of a prompt, a React interface that records the user's speech and saves it as a Blob, and a complex Python algorithm to score how well the human speech matches to the AI's.

Challenges we ran into

Some particular challenges we ran into was hooking up all the differing technologies together, especially between React and Python and sending an API call to ElevenLabs.

Accomplishments that we're proud of

We're proud of creating a tool that can easily be scalable to different prompts, different accents, and potentially even expanded into a Duolingo-type app.

What we learned

We learned how to use AI to expand human capabilities.

What's next for EchoStage

We want to push this app out further and perhaps create a Duolingo-style app where accent training is rewarded, so not only actors but normal people can use this app for quick bursts of education for their daily lives.

Built With

• elevenlabs
• javascript
• python
• react