BriefGPT.xyz
Apr, 2021
长尾实例分割的DropLoss
DropLoss for Long-Tail Instance Segmentation
HTML
PDF
Ting-I Hsieh, Esther Robb, Hwann-Tzong Chen, Jia-Bin Huang
TL;DR
本文提出一种新的自适应损失（DropLoss），用于解决目标检测中长尾类别分布下的损失不平衡问题，以改善罕见类别的预测结果，提高模型在LVIS数据集上的表现。
Abstract
long-tailed class distributions
are prevalent among the practical applications of object detection and
instance segmentation
. Prior work in long-tail
→