How to Pretrain YOLO11 with DINOv3 and LightyTrain

Transcript

So, Meta has just released a new Dino W3 self-supervised model. It is basically just for drawing bounding boxes, segmentation mask, depth maps, and all of that in a self-supervised way. So, you don't need the labels, all the ground truth data, drawing the bounding boxes and all that. We can actually use this as a foundation model both to train smal... Read More