1 paper · 1 filter
Martijn Oldenhof, Adam Arany, Yves Moreau +1
Training object detection models usually requires instance-level annotations, such as the positions and labels of all objects present in each image. Such supervision is unfortunate…