Skip to main content

Running AI Tasks with Hailo -With AI Kit

Introduction​

In the last chapter, we showed you how to set up the Raspberry Pi for various AI tasks. In this chapter, we will discuss how to perform object detection and pose estimation using the Hailo environment.

If you haven't set up your device yet, please follow the previous tutorial first and then return to this one.

The Hailo Model Zoo is a collection of pre-trained models using the COCO dataset for 80 classes. You can find various models trained by the Hailo team. In this tutorial, we will test YOLOv8, but you can explore other models, each with different architectures. The Hailo Model Zoo provides pre-trained models for high-performance deep learning applications.

Hailo provides different pre-trained models in ONNX/TF formats, as well as pre-compiled HEF (Hailo Executable Format) binary files to execute on Hailo devices.

Link to Model Zoo

In this tutorial, we will demonstrate object detection and pose estimation in the Hailo environment.

Object Detection​

  • Clone the repository:
git clone https://github.com/Seeed-Projects/Benchmarking-YOLOv8-on-Raspberry-PI-reComputer-r1000-and-AIkit-Hailo-8L.git
  • Navigate to directory
cd Benchmarking-YOLOv8-on-Raspberry-PI-reComputer-r1000-and-AIkit-Hailo-8L
  • Run object detection:
bash ./run.sh object-detection-hailo

object detection

We measured the inference speed of YOLOv8 for object detection with a 640×640 input resolution using the AI kit. With Hailo acceleration, it reached 29.5 FPS.

Pose Estimation​

  • Clone the repository (if not already):
git clone https://github.com/Seeed-Projects/Benchmarking-YOLOv8-on-Raspberry-PI-reComputer-r1000-and-AIkit-Hailo-8L.git
  • Navigate to directory
cd Benchmarking-YOLOv8-on-Raspberry-PI-reComputer-r1000-and-AIkit-Hailo-8L
  • Run object detection:
bash run.sh pose-estimation-hailo

pose estimation

The inference speed of YOLOv8 for pose estimation with a 640×640 input resolution using Hailo acceleration and the AI kit reached 27 FPS.