Adversarial Meta-Learning

2018/06/08 by Chengxiang Yin, Jian Tang, Yin, Chengxiang +5 · 30 citations
Computer Science · Engineering · Mathematics · #Advanced Neural Network Applications #Adversarial Robustness in Machine Learning #Adversarial system #Artificial intelligence #Computer science #Domain Adaptation and Few-Shot Learning #Engineering #FOS: Computer and information sciences #Initialization #Machine Learning (cs.LG) #Machine Learning (stat.ML) #Machine learning #Meta learning (computer science) #Robustness (evolution) #Task (project management) #cs.LG #stat.ML

paper · pdf · doi:10.48550/arxiv.1806.03316

published in arXiv (Cornell University) (Cornell University) · 11 pages

openalex publication_date 2018/06/08 · arxiv created 2020/06/20 · arxiv updated 2020/06/23 · openalex created_date 2025/10/10 · openalex updated_date 2026/08/08

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.

Citations

• Learning with a Strong Adversary
• Prototypical Networks for Few-shot Learning
• Meta Continual Learning
• Learning Multiple Layers of Features from Tiny Images

Cited by

• Doodle Your Keypoints: Sketch-Based Few-Shot Keypoint Detection
• Task-Robust Model-Agnostic Meta-Learning
• Long-term Cross Adversarial Training: A Robust Meta-learning Method for Few-shot Classification Tasks
• Adversarially Pretrained Transformers may be Universally Robust In-Context Learners
• Adversarially Robust Few-Shot Learning: A Meta-Learning Approach

Related

• On Fast Adversarial Robustness Adaptation in Model-Agnostic Meta-Learning
• Yet Meta Learning Can Adapt Fast, It Can Also Break Easily
• Learning Transferable Adversarial Robust Representations via Multi-view Consistency
• Long-term Cross Adversarial Training: A Robust Meta-learning Method for Few-shot Classification Tasks
• Adversarial Attacks are a Surprisingly Strong Baseline for Poisoning Few-Shot Meta-Learners