Hey,

My team's goal is to run yolov8 model with custom weights & labels we've trained, on a folder of grayscale images, whenever we run some python script, and save the outputs to a different folder.

We want to use the drone's gpu for the inference, so i went over the documents you guys provide, and i get that we need to make use of libmodal-pipe. However, when i try to compile the examples, i get linker errors.

It's VOXL2 SDK 1.4.1.
Would love some help, I can provide further logs etc. but I'd love to know if we could maybe take a different approach for our goal?

Thanks,
Itzik.