Colin Samplawski, Jannik Wolff, Tassilo Klein, Moin Nabi
TL;DR本篇论文研究了利用标签信息、属性以及关系图加强零样本学习和广义零样本学习,基于这个方法,本文在 CUB 和 SUN 数据集上取得了比强基线更好的结果。
Abstract
The task of zero-shot learning (ZSL) requires correctly predicting the label
of samples from classes which were unseen at training time. This is achieved by
leveraging side information about class labels, such as label