README.md

metadata

license: agpl-3.0
library_name: ultralytics
pipeline_tag: object-detection
tags:
  - yolo
  - ultralytics
  - yolov11
  - object-detection
  - ppe-detection
  - computer-vision
  - safety
datasets:
  - custom

PPE Detection with YOLOv11

This model is a specialized version of YOLOv11, fine-tuned to detect Personal Protective Equipment (PPE) in industrial and construction environments. It is designed to enhance occupational safety by monitoring the use of helmets and safety vests in real-time.

🚀 Quick Start (Usage)

You don't need to download the weights manually. You can load and run the model directly using the Python code below:

from ultralytics import YOLO
from huggingface_hub import hf_hub_download

# Downloading and loading the model
model_path = hf_hub_download(repo_id="melihuzunoglu/ppe-detection", filename="best.pt")
model = YOLO(model_path)

# Run inference
results = model.predict(source="construction_site.jpg", conf=0.25, save=True)

✅ Supported Classes (Labels)

The model follows a hierarchical detection strategy for higher accuracy:

Human: Detects the person/worker as the primary anchor.

Helmet: Detects safety helmets worn on the head.

No-Helmet: Specifically identifies heads without safety helmets.

Vest: Detects high-visibility safety vests.

Note: For "No-Vest" detection, the logic is based on the absence of a vest label within a detected human bounding box.

📊 Model Information

Architecture: YOLOv11 (Ultralytics)

Task: Object Detection (PPE Compliance)

Input Resolution: 640x640 pixels

Training Strategy: Hierarchical labeling (Human as anchor) to reduce false negatives in complex backgrounds.

🎯 Target Applications

Construction Sites: Real-time monitoring of helmet and vest compliance.

Industrial Plants: Automated safety audits for manufacturing floors.

Mining & Logistics: Ensuring worker visibility and protection in hazardous zones.

Safety Training: Providing visual feedback during safety drills.

🛠 Training Details

The model was trained using the Ultralytics framework. The dataset was meticulously curated and pre-processed via Roboflow, utilizing a hierarchical approach where equipment is detected in relation to the human figure. This ensures that the model focuses on the equipment's visual features while maintaining spatial awareness of the worker.

👤 Developer

Author: Melih Uzunoğlu Linkedin

Framework: Ultralytics YOLOv11

Dataset Source: Roboflow

Disclaimer

This model is developed for educational and research purposes. While it provides high-accuracy detections, it should be used as a supplementary tool alongside professional safety inspections in critical workplace environments.