Few-shot Learning 小样本学习在图像领域的研究现状( 五 )


2.2.2 Meta2.3基于优化的小样本学习( Based)
Based 方法认为普通的梯度下降方法难以在 few-shot 场景下拟合,因此通过调整优化方法来完成小样本分类的任务 。
as a model for few-shot 论文详解
[7] Ravi, , and Hugo . “ as a model for few-shot .” (2016).
Model- meta- for fastof deep 论文详解
[8] Finn, ,, and. “Model- meta- for fastof deep .”of the 34thon- 70. JMLR. org, 2017.