I. Tutorial Overview

YoloV13 is an object detection model proposed by a joint research team comprising Tsinghua University, Taiyuan University of Technology, Xi'an Jiaotong University, and other institutions in June 2025. Building upon the real-time detection advantages of the YOLO series, this model introduces new mechanisms such as hypergraph enhancement, high-order semantic modeling, and lightweight structural reconstruction. It achieves comprehensive leadership on mainstream datasets including MS COCO and Pascal VOC, demonstrating superior generalization capabilities and deployment practicality. The related paper can be found at YOLOv13: Real-Time Object Detection with Hypergraph-Enhanced Adaptive Visual Perception.

This tutorial uses a single GPU RTX 5090.

II. Project Examples

Image

Video

III. Execution Steps

1. After starting the container, click the API address to access the web interface.

2. Usage Instructions

If "Bad Gateway" is displayed, it indicates that the model is initializing. Due to its size, please wait approximately 2–3 minutes before refreshing the page.

Image

Video

Parameter Descriptions

• Model: yolov13n.pt (Nano), yolov13s.pt (Small), yolov13l.pt (Large), yolov13x.pt (Extra Large). Larger models offer higher accuracy (mAP) but require more parameters, computational cost (FLOPs), and inference time.
• Confidence Threshold: Confidence threshold.
• IoU Threshold: Intersection over Union threshold used for Non-Maximum Suppression (NMS).
• Max Detections Per Image: Maximum number of bounding boxes detected per image.

🖌️ If you come across excellent projects, feel free to leave a message recommending them! Additionally, we have established a tutorial exchange group; welcome friends to scan the QR code below, note "[SD Tutorials]" when joining, and discuss various technical issues or share application results ↓

[JIALIU_0]

Citation Information

The citation information for this project is as follows:

@article{yolov13,
  title={YOLOv13: Real-Time Object Detection with Hypergraph-Enhanced Adaptive Visual Perception},
  author={Lei, Mengqi and Li, Siqi and Wu, Yihong and et al.},
  journal={arXiv preprint arXiv:2506.17733},
  year={2025}
}