Any Object Tracking Tutorial on VisionFive2

Disclaimer: I am a newbie in this field so if i asked stupid questions please don’t be too harsh :cry:

I noticed there are very less resources on object tracking tutorial on VisionFive2, are there are tutorials from head to toes on the installation of frameworks such as PyTorch and Tensorflow, and the implementation of DL algorithm such as Yolo or CNN? Or are the installations and implementations similar to Raspberry Pi usage?

The reason I am asking this is because I read about something regarding PyTorch being incompatible with RISC-V. If you guys have any tutorial links or are interested to give me and overview on how i can do this, I will be extremely grateful! Thank You!

Hello selamatengineer, I’m very interested in your research, could you provide more information regarding how PyTorch incompatible with RISC-V? I am looking forward to any tutorial link too from the experts of this forum.

VF2’s CPU is JH7110 .
It is not have AI module and rv64gc is not include AI instruction(Is risc-v have AI instrucion? I remember it is no).
So Any AI engine run on standard risc-v is runing without accelerated.

You have two choose:

  1. Find out howto use common float-point unit to run those AI engine.
  2. Find out howto use vulkan to run AI. JH7110 have a GPU which degisned by IMG. They released a close source driver.