VFNet

2000년 도입 · 논문 8편에서 사용

VarifocalNet is a method aimed at accurately ranking a huge number of candidate detections in object detection. It consists of a new loss function, named Varifocal Loss, for training a dense object detector to predict the IACS, and a new efficient star-shaped bounding box feature representation for estimating the IACS and refining coarse bounding boxes. Combining these two new components and a bounding box refinement branch, results in a dense object detector on the FCOS architecture, what the authors call VarifocalNet or VFNet for short.

출처: VarifocalNet: An IoU-aware Dense Object Detector

소개 논문: VarifocalNet: An IoU-aware Dense Object Detector