MIRROR

What if your mirror could detect fatigue, stress, and recovery instantly? MIRROR uses AI to analyze your face and provide real-time wellness insights.

Comment

Inspiration

What if your mirror could detect early signs of fatigue, stress, and recovery — instantly?

We wanted to explore how AI could move from reactive healthcare to preventive insights. Your face carries subtle signals of your internal state, and we believed it was possible to capture those signals in real time using computer vision and AI.

This project was also inspired by the idea of making health feedback more intuitive and accessible — something as simple as looking into a mirror.

What it does

MIRROR is an AI-powered preventive health system that analyzes facial signals to detect early signs of fatigue, stress, and recovery.

Using a simple photo, the system:

  • Evaluates fatigue, stress, and hydration levels
  • Generates a wellness score
  • Provides real-time AI feedback (via Megan)
  • Tracks sessions and compares changes over time

It transforms a simple image into actionable health insights.

How we built it

We built MIRROR using:

  • Python + Streamlit for the interface
  • Google Gemini API for AI reasoning and feedback generation
  • Computer vision models for facial analysis
  • SQLite for session tracking and history

The system combines structured outputs (scores) with natural language AI feedback to create a human-like health assistant experience.

Challenges we ran into

One of the biggest challenges was balancing accuracy and interpretation.

Facial signals are subtle, and translating them into meaningful health indicators required careful design of prompts and outputs.

We also had to ensure:

  • Clean and structured AI responses
  • Real-time performance
  • A smooth and intuitive user experience

Accomplishments that we're proud of

We successfully built a fully functional AI system that:

  • Produces real-time health insights from a single image
  • Generates contextual, human-like feedback
  • Tracks and compares wellness over time

Most importantly, we created something that feels like a real product — not just a demo.

What we learned

We learned how to:

  • Combine computer vision with generative AI
  • Structure AI outputs reliably using prompt engineering
  • Build a full-stack AI product under time constraints

This project pushed us to think beyond models — and focus on real-world usability.

What's next for MIRROR

Next steps include:

  • Improving model accuracy with more data
  • Adding real-time video analysis
  • Personalizing insights over time
  • Integrating wearable or health data

Our vision is to turn MIRROR into a true preventive health assistant — one that helps people understand their body before problems arise

Built With

Share this project:

Updates