Adversarial Meta-Learning

TL;DR

This paper introduces ADML, a meta-learning algorithm that effectively utilizes both clean and adversarial samples to improve model initialization, achieving robustness and high accuracy even with limited or contaminated data.

Contribution

The paper presents ADML, a novel adversarial meta-learning algorithm that enhances robustness against adversarial samples and performs well with limited or contaminated data.

Findings

ADML outperforms existing meta-learning algorithms on MiniImageNet and CIFAR100.

ADML maintains high accuracy with adversarial samples, showing robustness.

Experimental results demonstrate ADML's effectiveness in adversarial and limited-data scenarios.

Abstract

Meta-learning enables a model to learn from very limited data to undertake a new task. In this paper, we study the general meta-learning with adversarial samples. We present a meta-learning algorithm, ADML (ADversarial Meta-Learner), which leverages clean and adversarial samples to optimize the initialization of a learning model in an adversarial manner. ADML leads to the following desirable properties: 1) it turns out to be very effective even in the cases with only clean samples; 2) it is robust to adversarial samples, i.e., unlike other meta-learning algorithms, it only leads to a minor performance degradation when there are adversarial samples; 3) it sheds light on tackling the cases with limited and even contaminated samples. It has been shown by extensive experimental results that ADML consistently outperforms three representative meta-learning algorithms in the cases involving adversarial samples, on two widely-used image datasets, MiniImageNet and CIFAR100, in terms of both accuracy and robustness.