Gemini Visual Assistant

• 

  architecture diagram

Inspiration

I have a deaf friend that is always thankful that she could see, and keeps saying that being blind is the real challenge as she can hear with hearing aid.

What it does

the app describes for the visually impaired people their surroundings.

How we built it

we used flutter to build the app, and firebase ai_logic as the backend

Challenges we ran into

making the live gemini api speak directly through the app, so we switched to flutter_tts to capture text and convert it to audio (speaking).

Accomplishments that we're proud of

this will be an open-source project to help build the visually impaired people better understand their surroundings.

What we learned

1. after closing our eyes to test the app, we learned that seeing is a bliss that we don't often be thankful for.
2. Gemini Live API is a powerful tool that can be used in different ways.

What's next for Gemini Visual Assistant

It's an MVP and open to contributions on github to achieve a final product that can be published in the Google Play Store.

Built With

• dart
• firebase
• firestore
• flutter
• gemini