Are we allowed to fine-tune an object detection model (e.g. YOLO) on the provided competition videos and then use that fine-tuned model only to detect/track vehicles and extract features?
Hey,
I believe that's one of the most natural things to do here. So my guess would be yes.
And honestly, you don't need any fine-tuning, there are plenty of pretrained models out there. You can make use of them. Cheers!!
Hey,
I believe that's one of the most natural things to do here. So my guess would be yes.
And honestly, you don't need any fine-tuning, there are plenty of pretrained models out there. You can make use of them. Cheers!!