1 paper
Zhizhong Chai, Huangjing Lin, Luyang Luo +2
Most of the existing object detection works are based on the bounding box annotation: each object has a precise annotated box. However, for rib fractures, the bounding box annotati…