Quickly Experience YOLOv10 on Jetson

May 31, 2024 · View on GitHub

Hello

💡 Here's an example of quickly deploying YOLOv10 on a Jetson device.

🔥 Hightlights:

  • Yolov10 is a state-of-the-art real-time object detection model. 🚀🔍
  • Jetson-examples is a toolkit designed to deploy containerized applications on NVIDIA Jetson devices. ✨
  • Jetson is powerful AI hardware platform for edge computing.💻

🛠️ Follow the tutorial below to quickly experience the performance of YOLOv10 on edge computing devices.

yolov10

Get a Jetson Orin Device 🛒

Device ModelDescriptionLink
Jetson Orin Nano Dev Kit, Orin Nano 8GB, 40TOPSDeveloper kit for NVIDIA Jetson Orin NanoBuy Here
reComputer J4012, powered by Orin NX 16GB, 100 TOPSEmbedded computer powered by Orin NXBuy Here

Getting Started

  • install jetson-examples by pip: 
    pip3 install jetson-examples
  • restart reComputer 
    sudo restart
  • run yolov10 on jetson in one line: 
    reComputer run yolov10
  • Please visit http://127.0.0.1:7860

Change Model

This example will automatically download the YOLOv10s model at startup. If you want to try different models, please use the following command to download the model and then select the appropriate model through the WebUI.

Note: You can also download the model via a browser and copy the model to /home/$USER/reComputer/yolov10/weights.

ModelDownload Command
YOLOv10-Nsudo wget -P /home/$USER/reComputer/yolov10/weights https://github.com/THU-MIG/yolov10/releases/download/v1.1/yolov10n.pt
YOLOv10-Ssudo wget -P /home/$USER/reComputer/yolov10/weights https://github.com/THU-MIG/yolov10/releases/download/v1.1/yolov10s.pt
YOLOv10-Msudo wget -P /home/$USER/reComputer/yolov10/weights https://github.com/THU-MIG/yolov10/releases/download/v1.1/yolov10m.pt
YOLOv10-Bsudo wget -P /home/$USER/reComputer/yolov10/weights https://github.com/THU-MIG/yolov10/releases/download/v1.1/yolov10b.pt
YOLOv10-Lsudo wget -P /home/$USER/reComputer/yolov10/weights https://github.com/THU-MIG/yolov10/releases/download/v1.1/yolov10l.pt
YOLOv10-Xsudo wget -P /home/$USER/reComputer/yolov10/weights https://github.com/THU-MIG/yolov10/releases/download/v1.1/yolov10x.pt

Build Docker Image

Our provided container is built based on the jetson-container. This example provide a Dockerfile, allowing you to build a more suitable container according to your needs.

sudo docker build -t yolov10-jetson .

Note: Additionally, you can train models, test models, and export models within the Docker container environment. For detailed information, please refer to THU-MIG/yolov10.

Reference