KLD

Learning High-Precision Bounding Box for Rotated Object Detection via Kullback-Leibler Divergence Introduction 本篇文章的四大亮点: Background Inductive Thinking of Loss Design: from Special Horizon to General