cross-domain few-shot learning (CDFSL) aims to acquire knowledge from limited
training data in the target domain by leveraging prior knowledge transferred
from source domains with abundant training samples. CDFSL faces challenges in
transferring knowledge across dissimilar domains and
本文提出了跨域 few-shot 学习的 Broader Study of Cross-Domain Few-Shot Learning (BSCD-FSL)基准,并对该基准的广泛实验表明了目前最先进的元学习方法被早期的元学习方法意外地超越,同时发现所有方法的准确性倾向于与数据集的相似性相关,这验证了该基准的价值,可指导未来的研究方向。