Cross-Domain Cross-Set Few-Shot Learning via Learning Compact and Aligned Representations

TL;DR

This paper introduces a novel method called stabPA for cross-domain cross-set few-shot learning, which learns compact and aligned representations to handle domain shifts and intra-class domain consistency, outperforming existing baselines.

Contribution

The paper proposes stabPA, a new approach that learns prototypical, compact, and cross-domain aligned representations specifically for the challenging CDSC-FSL setting, addressing both domain shift and intra-class domain consistency.

Findings

Outperforms multiple baselines significantly.

Achieves 6.0 points higher 5-shot accuracy on DomainNet.

Validates effectiveness on two new benchmarks.

Abstract

Few-shot learning (FSL) aims to recognize novel queries with only a few support samples through leveraging prior knowledge from a base dataset. In this paper, we consider the domain shift problem in FSL and aim to address the domain gap between the support set and the query set. Different from previous cross-domain FSL work (CD-FSL) that considers the domain shift between base and novel classes, the new problem, termed cross-domain cross-set FSL (CDSC-FSL), requires few-shot learners not only to adapt to the new domain, but also to be consistent between different domains within each novel class. To this end, we propose a novel approach, namely stabPA, to learn prototypical compact and cross-domain aligned representations, so that the domain shift and few-shot learning can be addressed simultaneously. We evaluate our approach on two new CDCS-FSL benchmarks built from the DomainNet and Office-Home datasets respectively. Remarkably, our approach outperforms multiple elaborated baselines by a large margin, e.g., improving 5-shot accuracy by 6.0 points on average on DomainNet. Code is available at https://github.com/WentaoChen0813/CDCS-FSL