---
license: mit
language:
- en
- zh
tags:
- yolov8
- tfjs
- hard-hat
- ultralytics
- yolo
- object-detection
library_name: ultralytics
library_version: 8.0.23
inference: false

datasets:
- keremberke/hard-hat-detection

model-index:
- name: keremberke/yolov8n-hard-hat-detection
  results:
  - task:
      type: object-detection

    dataset:
      type: keremberke/hard-hat-detection
      name: hard-hat-detection
      split: validation

    metrics:
      - type: precision  # since mAP@0.5 is not available on hf.co/metrics
        value: 0.83633  # min: 0.0 - max: 1.0
        name: mAP@0.5(box)
---

This model is built using tfjs and is based on the YOLOv8n architecture. It is capable of detecting two classes of objects: people wearing safety helmets and those who are not.

该模型使用tfjs构建，基于YOLOv8n架构，可以检测两类物体：戴安全帽的人和未戴安全帽的人。

This model is converted from https://huggingface.co/keremberke/yolov8n-hard-hat-detection

该模型转换自 https://huggingface.co/keremberke/yolov8n-hard-hat-detection

<div align="center">
  <img width="640" alt="keremberke/yolov8n-hard-hat-detection" src="https://huggingface.co/keremberke/yolov8n-hard-hat-detection/resolve/main/thumbnail.jpg">
</div>


### Supported Labels

```JSON
["Hardhat", "NO-Hardhat"]
```

### How to use

- Clone [this github repo](https://github.com/lanseria/yolov8-tfjs-vue-webrtc-demo)
- Read this repo readme