How can I write alls script for yolov8n_pose

General
1

I only understand it is doing some pre/post processing. But don’t know how to write our own. What are available options/syntax.

I am new to ML so If it ML specific can you guide me where to find information about it. I tried searching did not find anything similar.

2

Hey @saurabh

Hailo provides several resources that can help you understand and implement these processes:

  1. Hailo Application Code Examples:
    Check out GitHub - hailo-ai/Hailo-Application-Code-Examples for various examples of how we implement pre- and post-processing. This repository contains practical code samples that you can study and adapt for your needs.

  2. Raspberry Pi 5 Examples with GStreamer:
    If you’re working with Raspberry Pi 5 and want to use GStreamer, visit GitHub - hailo-ai/hailo-rpi5-examples. This resource provides examples specifically tailored for the Raspberry Pi 5 platform using GStreamer pipelines.

  3. Hailo Model Zoo:
    For additional examples and a deeper dive into post-processing, explore the Hailo Model Zoo at hailo_model_zoo/hailo_model_zoo/core/postprocessing at 918f9676b4e7bd5ed9e8be09684ff66c21188217 · hailo-ai/hailo_model_zoo · GitHub. This section of the Model Zoo contains various post-processing implementations that you can study and use as a reference.

These resources should give you a good starting point for understanding how to implement pre- and post-processing for YOLOv8n pose estimation in the Hailo ecosystem. They provide practical examples and code that you can examine to learn the available options and syntax.

If you have any specific questions about the implementations you find in these resources, feel free to ask, and I’ll be happy to provide further clarification or guidance.

3

Thank you. I am able to figure out. I copied the yolov8s_pose json and scripts and modified a little bit. End nodes are same. So working without any issue.

1 Like