跨域小样本---cross domain few shot---第一弹
cross domain Few-shot 存在问题域间类别不重合目标域时标签数据少存在域偏差A Broader Study of Cross-Domain Few-Shot Learning(ECCV 2020 link)提出了cross domain few-shot learning的标准,在miniImageNetminiImageNetminiImageNet上训练模型,随后采用N−way,K−shotN-way,K-shotN−way,K−shot的方式,迁移到目标数据集上。#.