IAN: Intelligent Accessibility Navigator

A real-time visual-assistance system for the visually impaired utilizing Gemini Vision capabilities, deployed on Google Cloud Run.

Gemini 2.5 VisionGoogle Cloud RunText-to-SpeechPythonREST API

Problem Statement

Visually impaired individuals face daily challenges navigating physical environments without real-time contextual information. Existing assistive tools often provide limited scene understanding. IAN leverages multimodal AI to provide natural language descriptions of visual scenes in real-time.

Architecture

Processing Pipeline

Camera InputReal-time video
Gemini Vision2.5 Flash analysis
Scene DescriptionNatural language
TTS OutputSpoken guidance

Visual Assets

Demo recording and accessibility testing screenshots coming soon.

Measurable Results

Real-Time

Scene description pipeline

Cloud Run

Serverless deployment

Multimodal

Vision + Language + Speech

Accessible

Designed for visually impaired users

What I Learned

  • Designing for accessibility-first rather than retrofitting — the entire architecture serves users with visual impairments.
  • Using Gemini Vision for real-time scene understanding and converting visual data to natural language.
  • Deploying latency-sensitive AI services on Cloud Run for auto-scaling serverless inference.
  • TTS integration challenges — balancing response speed with natural-sounding spoken output.

Zaynul Abedin Miah

CS Student Building AI & Cloud-Native Tools

BRAC University '27

Partner Resources

Check Dubbing v2 Access
Browse Cloud-Native Certifications

Code ZAYNUL for 30% off, if still valid.

Affiliate links. I may earn a commission if you sign up.

© 2026 Zaynul Abedin Miah. All rights reserved.

Trademarks and logos belong to their respective owners and are used only to identify public community affiliations.