Exploiting Style Transfer-based Task Augmentation for Cross-Domain Few-Shot Learning

TL;DR

This paper introduces TAML, a style transfer-based task augmentation method for cross-domain few-shot learning, enhancing model generalization across domains by increasing style diversity during training.

Contribution

It proposes novel style transfer and feature modulation techniques to improve domain generalization in few-shot learning, supported by theoretical analysis and extensive experiments.

Findings

TAML improves cross-domain generalization in few-shot learning.

Style augmentation increases training task diversity.

Method outperforms existing approaches on benchmark datasets.

Abstract

In cross-domain few-shot learning, the core issue is that the model trained on source domains struggles to generalize to the target domain, especially when the domain shift is large. Motivated by the observation that the domain shift between training tasks and target tasks usually can reflect in their style variation, we propose Task Augmented Meta-Learning (TAML) to conduct style transfer-based task augmentation to improve the domain generalization ability. Firstly, Multi-task Interpolation (MTI) is introduced to fuse features from multiple tasks with different styles, which makes more diverse styles available. Furthermore, a novel task-augmentation strategy called Multi-Task Style Transfer (MTST) is proposed to perform style transfer on existing tasks to learn discriminative style-independent features. We also introduce a Feature Modulation module (FM) to add random styles and improve generalization of the model. The proposed TAML increases the diversity of styles of training tasks, and contributes to training a model with better domain generalization ability. The effectiveness is demonstrated via theoretical analysis and thorough experiments on two popular cross-domain few-shot benchmarks.