ReFine: Re-randomization before Fine-tuning for Cross-domain Few-shot Learning

TL;DR

This paper introduces ReFine, a method that re-randomizes source-trained model parameters before fine-tuning, significantly enhancing cross-domain few-shot learning performance by mitigating domain differences.

Contribution

It proposes a novel re-randomization technique that improves transfer learning effectiveness in cross-domain few-shot scenarios.

Findings

ReFine improves few-shot learning accuracy across diverse domain shifts.

Re-randomization enhances fine-tuning efficiency on target domains.

The method outperforms existing transfer learning approaches in CD-FSL.

Abstract

Cross-domain few-shot learning (CD-FSL), where there are few target samples under extreme differences between source and target domains, has recently attracted huge attention. Recent studies on CD-FSL generally focus on transfer learning based approaches, where a neural network is pre-trained on popular labeled source domain datasets and then transferred to target domain data. Although the labeled datasets may provide suitable initial parameters for the target data, the domain difference between the source and target might hinder fine-tuning on the target domain. This paper proposes a simple yet powerful method that re-randomizes the parameters fitted on the source domain before adapting to the target data. The re-randomization resets source-specific parameters of the source pre-trained model and thus facilitates fine-tuning on the target domain, improving few-shot performance.