Adversarial Meta-Learning

By Chengxiang Yin, Jian Tang, Zhiyuan Xu, Yanzhi Wang. 2018

46 citations recorded by the source corpus.

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.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Optimization & Training

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Vision and Pattern Recognition

In Optimization & Training

Citations in the corpus

Repurposing Pretrained Models for Robust Out-of-domain Few-Shot Learning (2021, methodology): “ADversarialMeta-Learner (Yin et al., 2018) utilized one-step adversarial attack for generating adversarial samples during meta-training.”
Meta-Learning in Neural Networks: A Survey (2020, background): “Analogous to the case in domain-shift, an under-studied potential application of meta-learning is to train the learning algorithm end-to-end for robustness by deﬁning a meta-loss in terms of performance under adversarial attack [94], [283].”
A Collaborative Learning Framework via Federated Meta-Learning (2020, background): “It has been shown in [11] and [10] that meta-learning algorithms (e.g., MAML) are vulnerable to adversarial attacks, leading to possible signiﬁcant performance degradation of the locally fast adapted model at the target when facing perturbed data inputs.”
Robust Few-Shot Learning with Adversarially Queried Meta-Learners (2019, methodology): “ADML is a MAML variant, specifically designed for robustness, which employs adversarial training (Yin et al., 2018).”
Few-Shot Adversarial Learning of Realistic Neural Talking Head Models (2019, methodology): “Thus, data-augmentation GAN [2], Meta-GAN [43], adversarial meta-learning [41] use adversarially-trained networks to generate additional examples for classes unseen at the meta-learning stage.”
Hierarchical Graph Attention Network for Few-shot Visual-Semantic Learning (2021, background): “Currently, for general few-shot learning problems, meta-learning [29, 36, 52] has become a standard methodology.”
Adversarially Robust Few-Shot Learning: A Meta-Learning Approach (2019, background): “ADML is a MAML variant speciﬁcally designed speciﬁcally for robustness [33].”