YOLOv11 Environment Setup

[!IMPORTANT]

• The factory image has pre-configured the YOLOv11 environment. This tutorial is provided for users who need to set up the environment themselves. If you want to experience the complete functionality directly, you can skip this section!!!
• This tutorial is only for Jetson Orin and NX controllers, not applicable to Raspberry Pi or RDK X5.

0. YOLOv11 Introduction

1. YOLO (You Only Look Once) is a popular object detection and image segmentation model developed by Joseph Redmon and Ali Farhadi at the University of Washington. YOLO was launched in 2015 and has become widely popular for its high speed and accuracy.
2. Ultralytics' latest YOLO models deliver state-of-the-art (SOTA) performance across multiple tasks, including object detection, segmentation, pose estimation, tracking, and classification, and can be deployed in various AI applications and domains.
3. YOLOv11 is the latest version of the widely acclaimed real-time object detection and image segmentation model. Based on cutting-edge technologies in deep learning and computer vision, YOLO11 offers unparalleled performance in speed and accuracy. Its streamlined design makes it suitable for various applications and easily adaptable to different hardware platforms, from edge devices to cloud APIs.

YOLOv11 documentation link: https://docs.ultralytics.com

1. Course Content

1. Master the setup method for YOLOv11 runtime environment

2. Prerequisites

xxxxxxxxxx
sudo apt update
sudo apt install python3-pip -y
sudo pip install -U pip

3. Install Ultralytics

xxxxxxxxxx
pip install -U ultralytics

4. Configure Torch Environment

The factory image has downloaded related torch function installation packages. If you are using your own environment, you need to search and download the corresponding version from the PyTorch or NVIDIA official website pages. The process of setting up a DIY environment will not be explained here.

PyTorch official website: https://pytorch.org/get-started/locally/

View torch installation packages

xxxxxxxxxx
cd ~/software/torch_pkgs
ls

torch

xxxxxxxxxx
pip install ./torch-2.5.0a0+872d972e41.nv24.08-cp310-cp310-linux_aarch64.whl

torchvision

xxxxxxxxxx
pip install ./torchvision-0.20.0a0+afc54f7-cp310-cp310-linux_aarch64.whl

tensorflow

xxxxxxxxxx
pip install ./tensorflow-2.16.1+nv24.08-cp310-cp310-linux_aarch64.whl

5. Verify Installation

Verify Ultralytics

xxxxxxxxxx
python3 -c "import ultralytics; print(ultralytics.__version__)"

Verify Torch

xxxxxxxxxx
python3 -c "import torch; print(torch.__version__); print(torch.cuda.is_available())"

Verify Torchvision

xxxxxxxxxx
python3 -c "import torchvision; print(torchvision.__version__)"

6. Check System CUDA Version

The base image for Jetson series controllers is provided by NVIDIA official, and the GPU acceleration runtime environment is uniformly managed by the JetPack package. If you need to view related version information, you can use jtop.

xxxxxxxxxx
jtop

Click INFO to view version information