GeminEye

AI-Powered Vision: Guiding every step, Seeing Beyond Limits!!

Comment

🚀 Inspiration

Our inspiration was the need to benefit humanity, specifically targeting visually impaired individuals.
"What if we could give vision to those who cannot see?"
This question led us to build GeminEye—an AI-powered real-time navigation tool that enhances safety, inclusion, and independence for people with vision impairment.

🎯 What It Does

GeminEye is a real-time object detection application that:
✔️ Identifies objects & obstacles in the environment.
✔️ Provides navigation guidance with clear voice instructions.
✔️ Enables hands-free use with voice-controlled interaction.
✔️ Locates specific objects based on user commands (e.g., "Where is my backpack?").

🛠️ How We Built It

We used cutting-edge AI technology to make GeminEye fast, responsive, and accurate:

  • 🤖 AI Model: Gemini-1.5-Flash for scene analysis & guidance.
  • 🎥 OpenCV: Captures live video frames every 2 seconds for smooth AI feedback.
  • 🗣️ gTTS: Converts AI-generated text responses into speech output for accessibility.
  • 🔤 Python: The core programming language that ties it all together.

🛑 Challenges We Ran Into

1️⃣ Camera Access Issues in WSL

🔹 Initially, we couldn't open the camera using the WSL terminal.
🔹 We tested the camera with debug scripts and switched to Windows CMD to enable the live camera feature.

2️⃣ Live API Limitations

🔹 We couldn't integrate Gemini’s Live API for real-time streaming.
🔹 Instead, we captured frames every 2 seconds to simulate live video streaming without saving images to disk.

🏆 Accomplishments We're Proud Of

✅ Completed our first hackathon project!
✅ Overcame technical challenges like API integration & real-time video analysis.
✅ Built a working prototype within a short timeframe!

📚 What We Learned

🔹 Better API integration and real-time AI processing.
🔹 Optimizing AI prompts to get the most useful responses.
🔹 Persistence! We didn’t give up despite multiple technical challenges.

🔮 What's Next for GeminEye?

🚀 Scaling Up & Expanding Features

  • 📱 Smartphone & Wearable Integration: Making GeminEye accessible via smart glasses & mobile apps.
  • 🌍 Multilingual Support: Using Google Translate API to make navigation more inclusive globally.
  • 🗺️ GPS & Mapping: Integrating maps to enable self-navigation & location tracking.
  • 🟢 Live API Support: Using Gemini’s Live API for real-time, smoother analysis.
  • 🎨 Color Recognition for Colorblind Users: Helping users distinguish colors in their surroundings.

🔥 Why GeminEye Matters?

💡 AI should empower everyone—not just those who can see.
💡 GeminEye turns a smartphone into an AI-powered guide, making cities more inclusive, smarter, and safer for all.

Built With

Share this project:

Updates