Glance is a full-stack AI-powered web application designed to detect ships in Synthetic Aperture Radar (SAR) satellite imagery. Using a trained YOLOv8 deep learning model, users can upload satellite images and receive immediate, high-accuracy ship detections mapped onto an interactive interface.
- Framework: React 19 + Vite
- Language: TypeScript
- Styling: Tailwind CSS
- Visualizations: Cobe (interactive 3D globe)
- Database/Auth: Supabase JS
- Framework: FastAPI
- Server: Uvicorn
- AI/ML: PyTorch, Ultralytics (YOLOv8)
- Image Processing: OpenCV (headless)
/frontend- The React application (Vite, Tailwind, TypeScript)./backend- The FastAPI server and machine learning integration.Dockerfile- Containerization for the backend environment.requirements.txt- Python dependencies for the backend.
Follow the steps below to run both the backend server and the frontend application locally.
The backend handles image uploads, runs them through the trained YOLOv8 model, and returns the visualization and detection coordinates.
Prerequisites: Python 3.9+
# Navigate to the root directory
cd Glance_km
# Create and activate a virtual environment
python -m venv venv
source venv/bin/activate # On Windows use: venv\Scripts\activate
# Install the dependencies
pip install -r requirements.txt
# Run the FastAPI server
cd backend
uvicorn main:app --reload