---
license: cc-by-nc-4.0
datasets:
- Voxel51/fisheye8k
base_model:
- Ultralytics/YOLOv8
pipeline_tag: object-detection
tags:
- traffic
- urban
---

<p align="center">
  <img src="https://github.com/mkturkcan/yolo-fisheye/blob/main/assets/logo.png?raw=true"  width="180" />
</p>

<h3 align="center">
  <p>Circulaire: Real-Time Object Detection for Fisheye Cameras</p>
</h3>

<div align="center">
  <table>
    <tr>
      <th align="center"><strong>Architecture</strong></th>
      <th align="center"><strong>Recall</strong></th>
      <th align="center"><strong>mAP@50</strong></th>
    </tr>
    <tr>
      <td align="center">YOLOv0-pico</td>
      <td align="center">0.138</td>
      <td align="center">0.387</td>
    </tr>
    <tr>
      <td align="center">YOLOv8n</td>
      <td align="center">0.286</td>
      <td align="center">0.581</td>
    </tr>
    <tr>
      <td align="center">YOLOv11n</td>
      <td align="center">0.305</td>
      <td align="center">0.578</td>
    </tr>
    <tr>
      <td align="center">YOLOv11x</td>
      <td align="center"><strong>0.413</strong></td>
      <td align="center"><strong>0.634</strong></td>
    </tr>
    <tr>
      <td align="center">YOLO-Hyper</td>
      <td align="center"><strong>0.511</strong></td>
      <td align="center"><strong>0.711</strong></td>
    </tr>
  </table>
</div>