Dan Morris

We don't really have pre-trained weights in a state that are ready for sharing, but that's only out of cleanliness, not secrecy... email me at [email protected] and we'll figure out...

My two cents: as @biolars suggests, I've done a bunch of experimentation on this, and it's _almost_ always more trouble than it's worth. The core issue is that a quarter...

You won't have to twist my arm to get me to look at pictures of fox pups. :) If you have some images where "vanilla MD" struggles, or where your...

Lars sent me some images (~10k images from one camera), I had a blast trying a zillion different strategies for each of them (and looking at lots of fox pups)....

We don't do this right now, but you're right that since we have to read the images anyway, it wouldn't really slow processing down significantly (I think?), and I can...

Updating my response to this suggestion: rather than investing time in using the PyTorch data loader, I'd like to see someone experiment with YOLOv5's native inference tools (val.py and detect.py)...

Thanks for reporting this! This has been handled in the underlying library; see the issue that Peter linked to. That doesn't immediately fix the issue in AddaxAI; for unrelated reasons,...

I don't own the images so can't add them to the issue, but I added you to a relevant email thread, I'm sure the data owner will be happy to...

Thanks for the suggestions! I'm not trying to save time in this case, though, I'm just trying to make it through training without crashing during validation, which seems for whatever...

I'm currently testing with RF-DETR Nano, and training at an image size of 640px. Images are 1600px on the long edge. This is a test pass to get everything set...