(pronounced: ai·ris | aɪ.rɪs)

"AI That Opens Eyes"

AIris is a revolutionary wearable AI system that provides instant, contextual scene descriptions for visually impaired users. With a simple button press, users receive intelligent, real-time descriptions of their surroundings through advanced computer vision and natural language processing.

• Sub-2-second response time from capture to audio description
• Contextual intelligence with spatial awareness and safety prioritization
• Offline-first design with cloud enhancement capabilities
• Wearable form factor designed for comfort and accessibility
• Private audio delivery through integrated directional speakers

graph TB
    A[👓 Spectacle Camera] --> B[🖥️ Raspberry Pi 5]
    B --> C[🔊 Directional Speaker]
    B --> D[🔋 Portable Battery]
    B --> E[📱 Optional Phone Sync]
    
    style A fill:#4B4E9E,color:#fff
    style B fill:#C9AC78,color:#000
    style C fill:#4B4E9E,color:#fff

graph LR
    A[📷 Camera Interface] --> B[🧠 AI Engine]
    B --> C[🔊 Audio System]
    
    subgraph "AI Engine"
        D[🎯 Scene Analyzer]
        E[☁️ Groq API Client]
        F[🏠 Local Models]
    end
    
    subgraph "Audio System"
        G[🗣️ TTS Engine]
        H[🎵 Speaker Control]
    end
    
    style A fill:#E9E9E6
    style B fill:#4B4E9E,color:#fff
    style C fill:#E9E9E6

We're currently in the prototype and testing phase, working with a web interface to evaluate and optimize different multimodal AI models before hardware integration.

Our development team is using a local web interface to rapidly prototype and test various AI models:

We're currently testing and benchmarking multiple state-of-the-art vision-language models:

Model Evaluation

• Testing multiple vision-language models
• Benchmarking performance on Raspberry Pi 5
• Optimizing for speed vs. accuracy trade-offs

Web Interface Development

• Real-time model comparison dashboard
• Performance metrics visualization
• User experience prototyping

Performance Optimization

• Model quantization experiments
• Memory usage optimization
• Latency reduction techniques

• Custom hardware design and 3D modeling
• Wearable form factor development
• Field testing with target users

• ✅ Core software architecture
• ✅ AI model research and selection
• 🔄 Web interface development
• 🔄 Performance optimization
• ⏳ Audio system integration

• ⏳ Hardware design and 3D modeling
• ⏳ Wearable system integration
• ⏳ Field testing with users
• ⏳ Final optimization and documentation

This project will be developed by:

Name	Institution	ID	GitHub	Followers
Rajin Khan	North South University	2212708042
Saumik Saha Kabbya	North South University	2211204042

~ as part of CSE 499A/B at North South University, building upon the foundation of TapSense to advance accessibility technology.