Real-Time Object Detection Platform

A production-ready real-time object detection platform. It combines the power of YOLOv8 with a high-performance FastAPI backend and a modern React dashboard.

🚀 Key Features

• Live Stream Detection: Capture and analyze camera frames in real-time.
• Modern Dashboard: Glassmorphism UI with live metrics and latency tracking.
• Improved Accuracy: Powered by the YOLOv8 Medium model for high-precision detection (Phones, Pens, Laptops, etc.).
• Model Management: Dedicated tools for training, testing, and evaluating models.

📁 Project Structure

• frontend/: React (Vite) application with premium styling.
• backend/: FastAPI server logic and AI inference engine.
• model/: Scripts for training (train.py), testing (test.py), and evaluation (evaluate.py).
• dataset/: Scripts for downloading and augmenting training data.
• utils/: Core utilities like fps_counter.py, draw_boxes.py, and config.py.

📦 Installation

1. Backend Setup

cd backend
pip install -r requirements.txt
python app.py

2. Dataset & Model Setup (Optional)

# Download sample data
python dataset/downloader.py

# Test the model
python model/test.py

3. Frontend Setup

cd frontend
npm install
npm run dev

---

🛠️ Technology Stack

• Frontend: React, Lucide Icons, Axios.
• Backend: FastAPI, Ultralytics YOLOv8, OpenCV, Pillow.
• Inference: YOLOv8m (Medium) for balanced speed and accuracy.

---

🛡️ Original Implementation

This project is an original implementation designed for speed, modularity, and aesthetic excellence. It avoids legacy templates and focuses on modern, asynchronous web standards.