Inspiration

Many North American cities are not made for mainly humans in mind. They were made for cars so commuting and navigating through them is a challenge on foot.

This is especially challenging for the blind and visually impaired. Many cities lack sidewalks with tactile paving at intersections, or audible walking signals at traffic lights.

Echo(location) helps change that.

What it does

One has to simply open the site, point their camera, and take a picture. Echo will analyze the image and concisely return a summary, as well as the distance of the object to the user.

How we built it

Tech stack:

  • Expo
  • React native
  • GCP cloud functions
  • Elevenlabs text to speech

Challenges we ran into

Capturing the image and transforming it from base64, into binary data through a blob, transferring it over the network, then constructing the gemini prompt to accept it had me looking through a number of docs.

Initially the latency for Elevenlabs was a bit high (500-1000ms), but by choosing the right voice model I got it down to <100ms.

Accomplishments that I'm proud of

Making something functional and meanginful!

What's next for Echo

Show this to the visually impaired community and get their thoughts on it.

Built With

Share this project:

Updates