paper-with-me

홈 › Papers

Generative Poisoning Using Random Discriminators

2022-11-02 · Dirren van Vlijmen, Alex Kolmus, Zhuoran Liu, Zhengyu Zhao, Martha Larson

We introduce ShortcutGen, a new data poisoning attack that generates sample-dependent, error-minimizing perturbations by learning a generator. The key novelty of ShortcutGen is the use of a randomly-initialized discriminator, which provides spurious shortcuts needed for generating poisons. Different from recent, iterative methods, our ShortcutGen can generate perturbations with only one forward pass in a label-free manner, and compared to the only existing generative method, DeepConfuse, our ShortcutGen is faster and simpler to train while remaining competitive. We also demonstrate that integrating a simple augmentation strategy can further boost the robustness of ShortcutGen against early stopping, and combining augmentation and non-augmentation leads to new state-of-the-art results in terms of final validation accuracy, especially in the challenging, transfer scenario. Lastly, we speculate, through uncovering its working mechanism, that learning a more general representation space could allow ShortcutGen to work for unseen data.

📄 PDF Abstract BibTeX arXiv:2211.01086

Code (1)

liuzrcc/ImageShortcutSqueezing pytorch

Tasks

Data Poisoning

Similar Papers 제목 키워드 기반

Training generative networks using random discriminators

2019-04-22 · Babak Barazandeh, Meisam Razaviyayn, Maziar Sanjabi

In recent years, Generative Adversarial Networks (GANs) have drawn a lot of attentions for learning the underlying distribution of data in various applications. Despite their wide applicability, training GANs is notoriou…

FR-GAN: Fair and Robust Training

2019-09-25 · Yuji Roh, Kangwook Lee, Gyeong Jo Hwang, Steven Euijong Whang 외

We consider the problem of fair and robust model training in the presence of data poisoning. Ensuring fairness usually involves a tradeoff against accuracy, so if the data poisoning is mistakenly viewed as additional bia…

AttributeData PoisoningFairness

Stabilizing GAN Training with Multiple Random Projections

2017-05-22 · ICLR 2018 1 · Behnam Neyshabur, Srinadh Bhojanapalli, Ayan Chakrabarti

Training generative adversarial networks is unstable in high-dimensions as the true data distribution tends to be concentrated in a small fraction of the ambient space. The discriminator is then quickly able to classify …

DeepPoison: Feature Transfer Based Stealthy Poisoning Attack

2021-01-06 · Jinyin Chen, Longyuan Zhang, Haibin Zheng, Xueke Wang 외

Deep neural networks are susceptible to poisoning attacks by purposely polluted training data with specific triggers. As existing episodes mainly focused on attack success rate with patch-based samples, defense algorithm…

G-PATE: Scalable Differentially Private Data Generator via Private Aggregation of Teacher Discriminators

2019-06-21 · NeurIPS 2021 12 · Yunhui Long, Boxin Wang, Zhuolin Yang, Bhavya Kailkhura 외

Recent advances in machine learning have largely benefited from the massive accessible training data. However, large-scale data sharing has raised great privacy concerns. In this work, we propose a novel privacy-preservi…

BIG-bench Machine LearningPrivacy Preserving