1 paper
Yuqi Lin, Hengjia Li, Wenqi Shao +5
In this paper, we explore a principal way to enhance the quality of widely pre-existing coarse masks, enabling them to serve as reliable training data for segmentation models to re…