Large Margin Mechanism and Pseudo Query Set on Cross-Domain Few-Shot Learning

TL;DR

This paper introduces a large margin fine-tuning method with pseudo query set generation for cross-domain few-shot learning, significantly improving model adaptation across diverse datasets.

Contribution

It proposes a novel large margin fine-tuning approach using pseudo query images, enhancing cross-domain few-shot learning performance.

Findings

LMM-PQS outperforms baseline models significantly.

The method is robust across multiple diverse domains.

It effectively adapts pre-trained models with limited data.

Abstract

In recent years, few-shot learning problems have received a lot of attention. While methods in most previous works were trained and tested on datasets in one single domain, cross-domain few-shot learning is a brand-new branch of few-shot learning problems, where models handle datasets in different domains between training and testing phases. In this paper, to solve the problem that the model is pre-trained (meta-trained) on a single dataset while fine-tuned on datasets in four different domains, including common objects, satellite images, and medical images, we propose a novel large margin fine-tuning method (LMM-PQS), which generates pseudo query images from support images and fine-tunes the feature extraction modules with a large margin mechanism inspired by methods in face recognition. According to the experiment results, LMM-PQS surpasses the baseline models by a significant margin and demonstrates that our approach is robust and can easily adapt pre-trained models to new domains with few data.