Yusuke Kondo
Yusuke Kondo
Usually, Mask-R-CNN outputs many bounding boxes to an image. In your paper, the model use only ten bounding boxes for an image. How did you choose ten bounding boxes? (confidence,...
