Combining Domain-Specific Meta-Learners in the Parameter Space for Cross-Domain Few-Shot Classification

TL;DR

This paper introduces CosML, a meta-learning approach that combines domain-specific meta-learners in parameter space to improve cross-domain few-shot classification, demonstrating superior generalization to unseen domains.

Contribution

The paper proposes a novel optimization-based meta-learning method, CosML, which combines domain-specific meta-learners in parameter space for better cross-domain generalization.

Findings

CosML outperforms state-of-the-art methods in cross-domain few-shot classification.

CosML achieves strong generalization to unseen domains.

Combining domain-specific meta-learners improves adaptation to new tasks.

Abstract

The goal of few-shot classification is to learn a model that can classify novel classes using only a few training examples. Despite the promising results shown by existing meta-learning algorithms in solving the few-shot classification problem, there still remains an important challenge: how to generalize to unseen domains while meta-learning on multiple seen domains? In this paper, we propose an optimization-based meta-learning method, called Combining Domain-Specific Meta-Learners (CosML), that addresses the cross-domain few-shot classification problem. CosML first trains a set of meta-learners, one for each training domain, to learn prior knowledge (i.e., meta-parameters) specific to each domain. The domain-specific meta-learners are then combined in the \emph{parameter space}, by taking a weighted average of their meta-parameters, which is used as the initialization parameters of a task network that is quickly adapted to novel few-shot classification tasks in an unseen domain. Our experiments show that CosML outperforms a range of state-of-the-art methods and achieves strong cross-domain generalization ability.