Inspiration

When we visit a museum and enjoy the experience, we realize that museums are designed primarily for people who can see. There are many written narrations and visuals to experience, and this is how culture is transferred from generation to generation. So, how can blind people enjoy these cultural experiences too?

We were inspired by the many apps that try to help blind people, but we found that they still require a lot of assistance from other people, and the AI for image processing isn't really designed with their needs in mind. In conclusion, why don't we design an app specifically for blind people to experience and connect with these cultures?

What it does

We created "NaraNetra" to address the profound frustration and isolation visually impaired individuals experience in cultural spaces, which are overwhelmingly designed for sight.

This mobile application has two key features.

  1. Guidance Mode: The first feature is an AI-powered guide that assists users as they walk. It provides immediate and simple responses using four essential commands for clear direction: "Safe," "Turn Left," "Turn Right," and "Stop."
  2. Narration Mode: The second feature provides audio narration for blind users. The AI can describe visual information in the user's environment. For example, it can generate a detailed narration of art pieces like sculptures or paintings in a museum, making cultural experiences more accessible.

How we built it

  • First we discussed on what track to choose and to build.
  • And then, we consult our initial ideas to mentor, and from what we've got, we enhance the idea.
  • We then implemented it using next.js

Challenges we ran into

Latency so bad at first, it takes up to more than 10 seconds just to check one single image.

Accomplishments that we're proud of

  • Auto check environment condition using phone camera.
  • Reduce latency when detecting obstacles, via gemini vision.

What we learned

I think for the product part, we learn from the mentors that a good product doesn't have to own major features, instead we can just have one core features, but ensure it is impactful to our entire product.

What's next for NaraNetra

We would like to improve our accuracy on detecting obstacles and of course increase the safety of the user.

Built With

  • next.js
Share this project:

Updates