AuraVision AI

Multimodal AI vision & audio assistant empowering visually impaired users with real-time scene perception, OCR & voice Q&A powered by Gemini 1.5 Flash.

Comment

Inspiration

What it does

How we built it

Challenges we ran into

💡 Inspiration

Over 285 million visually impaired individuals face daily challenges with spatial navigation, reading printed documents, and identifying currency. Existing tools are fragmented, slow, and lack conversational context. AuraVision AI was created to provide an all-in-one, eyes-free multimodal vision assistant that acts as a real-time visual companion.

🚀 What it does

AuraVision AI turns any smartphone into an intelligent visual and audio companion:

  • Sub-Second Scene Perception: Instant spatial mapping and hazard narration powered by Google Gemini 1.5 Flash.
  • Ultra-HD Document & Prescription OCR: Accurately reads printed books, letters, and prescription labels with crystal-clear voice narration.
  • Instant Currency Detector: Identifies banknotes and denominations in real time.
  • Conversational Vision Memory: Enables blind users to ask voice follow-up questions referencing previous visual context without re-scanning.
  • Eyes-Free Blind UX: Full-screen double-tap scanning, voice commands, and tactile haptic feedback designed for non-visual usability.
  • Samsung Galaxy Foldable Adaptive Layout: Multi-window and foldable-ready responsive architecture.

🛠️ How we built it

  • Frontend: Flutter & Dart with zero-visual accessibility semantics and high-contrast styling.
  • AI Core: Google Gemini 1.5 Flash API with optimized low-latency streaming and multimodal buffers.
  • Monetization Engine: RevenueCat SDK powering flexible subscription tiers, paywalls, and judge evaluation code redemption.

🎁 Judge Evaluation Code

Hackathon judges can unlock full lifetime Pro features directly inside the app using the promo code: SHIPATON2026

🧗 Challenges we overcame

  1. Zero-Visual Interaction: Designing a seamless UX where users never need to aim for tiny buttons—solved via tactile haptic cues and global double-tap triggers.
  2. Real-time Latency: Optimizing image compression before calling Gemini 1.5 Flash to ensure voice feedback within 1-2 seconds.

🏅 Accomplishments that we're proud of

  • Production-ready release APK/AAB builds deployed.
  • Seamless multimodal vision + conversational memory integration.
  • Robust RevenueCat monetization architecture with promo code support.

🔮 What's next for AuraVision AI

  • Wearable integration for smart glasses and Galaxy Watch audio streams.
  • Offline edge-AI obstacle detection for low-connectivity environments.
  • Multi-language audio localization for global accessibility. ## Accomplishments that we're proud of

What we learned

What's next for AuraVision AI

Built With

  • accessibility
  • android
  • computer
  • dart
  • flutter
  • gemini-1.5-flash
  • google-gemini-api
  • multimodal-ai
  • ocr
  • revenuecat
Share this project:

Updates

Submission history