🔫 Real-Time Weapon Detection using YOLOv8

AI that spots weapons in real time. Just YOLOv8, a camera, and a model trained to protect.

This project harnesses the power of YOLOv8 to detect weapons like guns and knives directly from live webcam feeds or video footage. You can train it on a custom dataset from Kaggle or run it instantly using a pretrained best.pt model. Perfect for smart surveillance, AI-powered security systems, and anyone looking to fuse deep learning with real-world safety.

---

📁 Project Structure

weapon-detection/
├── dataset/ # YOLO-formatted dataset (downloaded from Kaggle)
├── best.pt # Pretrained model (optional)
├── detect.py # Real-time detection script
├── train_colab.ipynb # Google Colab training script
└── README.md

✅ Requirements

• Python 3.8+
• Ultralytics YOLOv8
• OpenCV
• PyTorch

🚀 Getting Started

Option 1: 🔽 Use Pretrained Model

1. Download best.pt from this GitHub repo or from your training.
2. Place it in your project directory.
3. Update detect.py:

   model_path = r"C:/..."

Option 2: 🧠 Train Your Own Model (Google Colab)

1. Open train_colab.ipynb in Colab.

2. Follow steps to:

   • Run training
   • Download best.pt after training

   from google.colab import files
   files.download('runs/detect/train/weights/best.pt')

3. Update detect.py:

   model_path = r"C:/..."

🔍 Real-Time Detection(Webcam)

Run the detect.py:

python detect.py

🤝 Open to Collaborate

Whether you're a developer, researcher, or enthusiast — if you're passionate about AI, computer vision, or building for public safety, let's connect!
Feel free to open issues, suggest features, or submit pull requests.
Always open to ideas, improvements, and partnerships.