Hi all,

My question is related to the execution of the code. I am currently running an object detection with usb camera on my laptop. I will transfer the code to make it work on a raspberry Pi instead of the computer. I just need to know if this will result in any change or downgrade in the performance of the code. Are the operations executed only with the VPU of the camera ? or if I change the host laptop to the camera , there will be a downgrade in results ?

Thank you in advance

    Hi MahdiMseddi ,
    Yes, all the computation will be done on the camera itself, including AI, CV, video encoding and stereo pipeline. Object detection performance on laptop will of course depend on your laptop, but below you can see what performance you can expect with OAK devices:
    https://docs.luxonis.com/projects/hardware/en/latest/pages/rvc/rvc2.html#rvc2-nn-performance
    I hope this helps!

      erik
      thank you Erik for your reply.
      "Object detection performance on laptop will of course depend on your laptop" How is this possible if all computation is done on the camera itself ?

      • erik replied to this.

        MahdiMseddi Sorry, I meant that by using webcam + laptop for inference. When using OAK, performance can be seen on above link.