Skip to main content

YOLOv8 Pose Estimation on reComputer R1000 with Hailo-8L

Introduction

YOLOv8 (You Only Look Once version 8) is the popular most YOLO series of real-time pose estimation models. It builds upon the strengths of its predecessors by introducing several advancements in speed, accuracy, and flexibility. The Raspberry-pi-AI-kit is used to accelerate inference speed, featuring a 13 tera-operations per second (TOPS) neural network inference accelerator built around the Hailo-8L chip.

This wiki demonstrates pose estimation using YOLOv8 on reComputer R1000 with and without Raspberry-pi-AI-kit acceleration. The Raspberry Pi AI Kit enhances the performance of the Raspberry Pi and unlock its potential in artificial intelligence and machine learning applications, like smart retail, smart traffic and more. Although the Raspberry AI Kit is designed for Raspberry Pi 5, we have experimented it on our CM4-powered edge gateway. Excited about turning our edge device into an intelligent IoT gateway!

Prepare Hardware

reComputer r1000Raspberry Pi AI Kit

Run this project

Step 1: Install AI kit

pir

Step 2: Update system & Set pcie to gen3

Updata system

Open terminal on the reCompuer R1000, and input command as follows to update your system.

sudo apt update
sudo apt full-upgrade

Set pcie to gen3

Open terminal on the reCompuer R1000, and input command as follows to config reCompuer R1000.

sudo raspi-config

Select option "6 Advanced Options"

pir

Then select option "A8 PCIe Speed"

pir

Choose "Yes" to enable PCIe Gen 3 mode

pir

Click "Finish" to exit

pir

Step 3: Install Hailo Software & Verify Installation

Install Hailo Software

Open terminal on the reCompuer R1000, and input command as follows to install Hailo software.

sudo apt install hailo-all
sudo reboot

Check Software and Hardware

Open terminal on the reCompuer R1000, and input command as follows to check if hailo-all have been installed.

hailortcli fw-control identify

The right result show as bellow:

pir

Open terminal on the reCompuer R1000, and input command as follows to check if hailo-8L have been connected.

lspci | grep Hailo

The right result show as bellow:

pir

Run YOLOv8

Open terminal on the reCompuer R1000, and input command as follows to run YOLOv8.

git clone https://github.com/Seeed-Projects/Benchmarking-YOLOv8-on-Raspberry-PI-reComputer-r1000-and-AIkit-Hailo-8L.git
cd Benchmarking-YOLOv8-on-Raspberry-PI-reComputer-r1000-and-AIkit-Hailo-8L
bash ./run.sh pose-estimation-hailo

Result

pir

Result

We compared the inference speed of YOLOv8 for pose estimation with input 640*640 resolution before and after acceleration using the AI kit. The results show that prior to acceleration, the inference speed was only 0.5 FPS, whereas after acceleration, it reached 27 FPS.

Project Outlook

In this project, we benchmark the running speed of YOLOv8 on pose estimation with and without AI kit. The result shows that the AI kit can greatly improve the performance of the edge device. And in the future, we will benchmark the running speed of YOLOv8 in different scenarios.

Tech Support & Product Discussion

Thank you for choosing our products! We are here to provide you with different support to ensure that your experience with our products is as smooth as possible. We offer several communication channels to cater to different preferences and needs.

Loading Comments...