| --- |
| language: |
| - en |
| license: mit |
| library_name: ultralytics |
| tags: |
| - object-detection |
| - yolo |
| - yolov11 |
| - computer-vision |
| - construction |
| - safety |
| - ppe |
| datasets: |
| - roboflow/construction-site-safety |
| metrics: |
| - mAP50 |
| --- |
| |
| # ConstructIQ Monitor - Construction Safety Detection (YOLOv11s) |
|
|
| This is a fine-tuned **YOLOv11s** object detection model designed specifically for construction site monitoring and safety compliance. It detects workers, heavy machinery, and Personal Protective Equipment (PPE) to automate hazard identification and site intelligence. |
|
|
| This model is a core component of the **[SiteSpectra / ConstructIQ](https://github.com/2005legend/ConstructIQ-Monitor)** computer vision pipeline. |
|
|
| ## 🏗️ Supported Classes |
| The model detects 10 distinct classes relevant to construction reality-capture: |
| - `Hardhat` |
| - `Mask` |
| - `NO-Hardhat` |
| - `NO-Mask` |
| - `NO-Safety Vest` |
| - `Person` |
| - `Safety Cone` |
| - `Safety Vest` |
| - `machinery` |
| - `vehicle` |
|
|
| ## 📊 Training & Validation Metrics |
| The model was fine-tuned for 20 epochs on a dataset of over 2,800 construction site images. It achieved strong validation metrics, particularly on critical safety and equipment classes: |
| - **Overall mAP50:** 80.6% |
| - **Machinery mAP50:** 92.0% |
| - **Hardhat mAP50:** 88.9% |
| - **Safety Vest mAP50:** 87.0% |
| - **Person mAP50:** 83.5% |
|
|
| ## 🚀 How to Use |
|
|
| You can easily use this model in Python using the `ultralytics` library. |
|
|
| ### Installation |
| ```bash |
| pip install ultralytics huggingface_hub |
| |
