johnosborne/splay-network-tracking-detection

Energy-Efficient Wildlife Tracking Using a Splay Tree Inspired Clustering Policy and Vision-Aided Analytics

This project implements a cohesive system that combines Splay Network Algorithms for efficient sensor communication with Real-Time AI Computer Vision for wildlife tracking. It is designed to simulate a high-tech conservation network where "Edge Nodes" (sensors) capture low-resolution data, and a central "Server" restores and analyzes it.

Key Features

1. Splay Network Simulation (The Backbone)

• —Dynamic Clustering: Visualizes how sensor nodes self-organize into clusters to save energy.
• —Splay-Tree Optimization: Uses splay algorithms to determine optimal routing paths to the gateway.
• —Energy Modeling: Simulates battery drain (green -> yellow -> red -> dead) and tracks network lifespan.
• —Live Dashboard: Monitors network health, tracking active vs. dead nodes and downtime.

2. AI Wildlife Analytics (The Intelligence)

• —Simulated Edge-to-Server Pipeline:
• —Edge Step: Takes raw video frames and downscales them to 128px "thumbnails" (mimicking low-bandwidth transmission).
• —Restoration Step: Uses SwinIR (simulated) on the server to upconvert thumbnails back to 640px.
• —Object Detection: Utilizing YOLOv8 (Nano) to identify animals in the restored feed.
• —Tracking: Implementing DeepSORT to assign unique IDs to animals and track their movement across frames.

How to Run

Option 1: Live Demo (Hugging Face)

This project is deployed on Hugging Face Spaces for high-performance AI inference. **View Live Demo**

Option 2: Run Locally

1. 1.Clone the Repository:

    git clone https://github.com/john-osborne-j/-splay-network.git
    cd splay-network

1. 1.Install Dependencies:

    python -m venv venv
    venv\Scripts\activate  # Windows
    pip install -r requirements.txt

1. 1.Run the App:

    python app.py

1. 1.Access: Open http://localhost:5000 in your browser.

🛠 Tech Stack

• —Backend: Python, Flask, Gunicorn
• —Frontend: HTML5, Vanilla CSS (Glassmorphism), JavaScript (Canvas API)
• —AI/ML: PyTorch, Ultralytics YOLOv8, DeepSORT, OpenCV
• —Deployment: Docker, Hugging Face Spaces

Project Structure

• —app.py: Main Flask server endpoints.
• —simulation.py: Logic for Splay Network nodes, energy, and routing.
• —detection.py: AI pipeline (SwinIR Restoration + YOLO Detection + DeepSORT).
• —static/: CSS styling and JavaScript simulation loop.
• —templates/: HTML frontend interface.

Created by John Osborne And Nishanth A Powered by Splay Algorithms & Modern AI