.png){kind=icon}
返回
A3T: accuracy aware adversarial training
.png){kind=icon}
OA
AI
DOI：10.1007/s10994-023-06341-w.png)
.png){kind=icon}
原文求助.png){kind=icon}
分享.png){kind=icon}
收藏
摘要
En 中文
Adversarial training has been empirically shown to be more prone to overfitting than standard training. The exact underlying reasons are still not fully understood. In this paper, we identify one cause of overfitting related to current practices of generating adversarial examples from misclassified samples. We show that, following current practice, adversarial examples from misclassified samples results in harder-to-classify samples than the original ones. This leads to a complex adjustment of the decision boundary during training and hence overfitting. To mitigate this issue, we propose A3T, an accuracy aware AT method that generate adversarial example differently for misclassified and correctly classified samples. We show that our approach achieves better generalization while maintaining comparable robustness to state-of-the-art AT methods on a wide range of computer vision, natural language processing, and tabular tasks.
Keyword:
Adversarial training
Overfitting in adversarial training
Accuracy aware adversarial training
期刊
IF:
2.9
论文数:
2.7K
被引数:
3.4W
机构
引用论文
Enhanced Reactivity of Dinuclear Copper(I) Acetylides in Dipolar Cycloadditions偶极环加成反应中双核乙炔铜 (I) 的反应性增强
PREAI
分享收藏Investigation of Ti mesh-supported anodes for direct borohydride fuel cells直接硼氢化物燃料电池用钛网支撑阳极的研究
PREAI
分享收藏Brain-derived neurotrophic factor as a model system for examining gene by environment interactions across development.png)
Neuroscience
OAAI
NeuroscienceIF0
OAAI分享收藏分享收藏Poultry Shelf-Life Enhancing Potential of Nanofibers and Nanoparticles Containing Porphyra dioica Extracts含有紫菜提取物的纳米纤维和纳米颗粒的家禽货架期延长潜力CoatingsOAAI
CoatingsIF0
OAAI分享收藏分享收藏
查看更多内容