# Better Safe Than Sorry: Preventing Delusive Adversaries with Adversarial Training

arXiv: https://arxiv.org/abs/2102.04716 · HTML: https://mlsanity.com/paper/2102.04716/

## Key facts

- Cited by 59 papers in the map (data through 14 Oct 2026).
- Builds on 84 of its 148 references that are in the map.
- First posted on arXiv 9 Feb 2021.

## Builds on (most cited first, 30 of 84)

- [Deep Residual Learning for Image Recognition](https://mlsanity.com/paper/1512.03385/) (2015)
- [Very Deep Convolutional Networks for Large-Scale Image Recognition](https://mlsanity.com/paper/1409.1556/) (2014)
- [A Simple Framework for Contrastive Learning of Visual Representations](https://mlsanity.com/paper/2002.05709/) (2020)
- [ImageNet Large Scale Visual Recognition Challenge](https://mlsanity.com/paper/1409.0575/) (2014)
- [Explaining and Harnessing Adversarial Examples](https://mlsanity.com/paper/1412.6572/) (2014)
- [Densely Connected Convolutional Networks](https://mlsanity.com/paper/1608.06993/) (2016)
- [Towards Deep Learning Models Resistant to Adversarial Attacks](https://mlsanity.com/paper/1706.06083/) (2017)
- [Intriguing properties of neural networks](https://mlsanity.com/paper/1312.6199/) (2013)
- [Understanding Black-box Predictions via Influence Functions](https://mlsanity.com/paper/1703.04730/) (2017)
- [Unsupervised Representation Learning by Predicting Image Rotations](https://mlsanity.com/paper/1803.07728/) (2018)
- [Theoretically Principled Trade-off between Robustness and Accuracy](https://mlsanity.com/paper/1901.08573/) (2019)
- [Obfuscated Gradients Give a False Sense of Security: Circumventing Defenses to Adversarial Examples](https://mlsanity.com/paper/1802.00420/) (2018)
- [Targeted Backdoor Attacks on Deep Learning Systems Using Data Poisoning](https://mlsanity.com/paper/1712.05526/) (2017)
- [Adversarial Examples Are Not Bugs, They Are Features](https://mlsanity.com/paper/1905.02175/) (2019)
- [Robustness May Be at Odds with Accuracy](https://mlsanity.com/paper/1805.12152/) (2018)
- [Universal adversarial perturbations](https://mlsanity.com/paper/1610.08401/) (2016)
- [Evasion Attacks against Machine Learning at Test Time](https://mlsanity.com/paper/1708.06131/) (2017)
- [Distillation as a Defense to Adversarial Perturbations against Deep Neural Networks](https://mlsanity.com/paper/1511.04508/) (2015)
- [RobustBench: a standardized adversarial robustness benchmark](https://mlsanity.com/paper/2010.09670/) (2020)
- [Self-supervised Visual Feature Learning with Deep Neural Networks: A Survey](https://mlsanity.com/paper/1902.06162/) (2019)
- [Provable defenses against adversarial examples via the convex outer adversarial polytope](https://mlsanity.com/paper/1711.00851/) (2017)
- [Poison Frogs! Targeted Clean-Label Poisoning Attacks on Neural Networks](https://mlsanity.com/paper/1804.00792/) (2018)
- [Overfitting in adversarially robust deep learning](https://mlsanity.com/paper/2002.11569/) (2020)
- [Spectral Signatures in Backdoor Attacks](https://mlsanity.com/paper/1811.00636/) (2018)
- [STRIP: A Defence Against Trojan Attacks on Deep Neural Networks](https://mlsanity.com/paper/1902.06531/) (2019)
- [Detecting Backdoor Attacks on Deep Neural Networks by Activation Clustering](https://mlsanity.com/paper/1811.03728/) (2018)
- [Wild Patterns: Ten Years After the Rise of Adversarial Machine Learning](https://mlsanity.com/paper/1712.03141/) (2017)
- [The Pitfalls of Simplicity Bias in Neural Networks](https://mlsanity.com/paper/2006.07710/) (2020)
- [Reflection Backdoor: A Natural Backdoor Attack on Deep Neural Networks](https://mlsanity.com/paper/2007.02343/) (2020)
- [Neural Attention Distillation: Erasing Backdoor Triggers from Deep Neural Networks](https://mlsanity.com/paper/2101.05930/) (2021)

## Built on by (most cited first, 30 of 59)

- [Adversarial Neuron Pruning Purifies Backdoored Deep Models](https://mlsanity.com/paper/2110.14430/) (2021)
- [Trustworthy AI: A Computational Perspective](https://mlsanity.com/paper/2107.06641/) (2021)
- [Autoregressive Perturbations for Data Poisoning](https://mlsanity.com/paper/2206.03693/) (2022)
- [MetaCloak: Preventing Unauthorized Subject-driven Text-to-image Diffusion-based Synthesis via Meta-learning](https://mlsanity.com/paper/2311.13127/) (2023)
- [Image Shortcut Squeezing: Countering Perturbative Availability Poisons with Compression](https://mlsanity.com/paper/2301.13838/) (2023)
- [Unlearnable Examples for Diffusion Models: Protect Data from Unauthorized Exploitation](https://mlsanity.com/paper/2306.01902/) (2023)
- [Robust Unlearnable Examples: Protecting Data Against Adversarial Learning](https://mlsanity.com/paper/2203.14533/) (2022)
- [Indiscriminate Poisoning Attacks on Unsupervised Contrastive Learning](https://mlsanity.com/paper/2202.11202/) (2022)
- [Not All Poisons are Created Equal: Robust Training against Data Poisoning](https://mlsanity.com/paper/2210.09671/) (2022)
- [Learning the Unlearnable: Adversarial Augmentations Suppress Unlearnable Example Attacks](https://mlsanity.com/paper/2303.15127/) (2023)
- [CUDA: Convolution-based Unlearnable Datasets](https://mlsanity.com/paper/2303.04278/) (2023)
- [Stable Unlearnable Example: Enhancing the Robustness of Unlearnable Examples via Stable Error-Minimizing Noise](https://mlsanity.com/paper/2311.13091/) (2023)
- [What Doesn't Kill You Makes You Robust(er): How to Adversarially Train against Data Poisoning](https://mlsanity.com/paper/2102.13624/) (2021)
- [Friendly Noise against Adversarial Noise: A Powerful Defense against Data Poisoning Attacks](https://mlsanity.com/paper/2208.10224/) (2022)
- [What Can We Learn from Unlearnable Datasets?](https://mlsanity.com/paper/2305.19254/) (2023)
- [The Devil's Advocate: Shattering the Illusion of Unexploitable Data using Diffusion Models](https://mlsanity.com/paper/2303.08500/) (2023)
- [Purify Unlearnable Examples via Rate-Constrained Variational Autoencoders](https://mlsanity.com/paper/2405.01460/) (2024)
- [Provably Unlearnable Data Examples](https://mlsanity.com/paper/2405.03316/) (2024)
- [On the Effectiveness of Adversarial Training against Backdoor Attacks](https://mlsanity.com/paper/2202.10627/) (2022)
- [Average of Pruning: Improving Performance and Stability of Out-of-Distribution Detection](https://mlsanity.com/paper/2303.01201/) (2023)
- [Fooling Adversarial Training with Inducing Noise](https://mlsanity.com/paper/2111.10130/) (2021)
- [Detection and Defense of Unlearnable Examples](https://mlsanity.com/paper/2312.08898/) (2023)
- [Semantic Deep Hiding for Robust Unlearnable Examples](https://mlsanity.com/paper/2406.17349/) (2024)
- [Can Adversarial Training Be Manipulated By Non-Robust Features?](https://mlsanity.com/paper/2201.13329/) (2022)
- [Game-Theoretic Unlearnable Example Generator](https://mlsanity.com/paper/2401.17523/) (2024)
- [Efficient Availability Attacks against Supervised and Contrastive Learning Simultaneously](https://mlsanity.com/paper/2402.04010/) (2024)
- [Sharpness-Aware Data Poisoning Attack](https://mlsanity.com/paper/2305.14851/) (2023)
- [Game-Theoretic Defenses for Robust Conformal Prediction Against Adversarial Attacks in Medical Imaging](https://mlsanity.com/paper/2411.04376/) (2024)
- [ECLIPSE: Expunging Clean-label Indiscriminate Poisons via Sparse Diffusion Purification](https://mlsanity.com/paper/2406.15093/) (2024)
- [APBench: A Unified Benchmark for Availability Poisoning Attacks and Defenses](https://mlsanity.com/paper/2308.03258/) (2023)

## Abstract (arXiv)

Delusive attacks aim to substantially deteriorate the test accuracy of the learning model by slightly perturbing the features of correctly labeled training examples. By formalizing this malicious attack as finding the worst-case training data within a specific ∞-Wasserstein ball, we show that minimizing adversarial risk on the perturbed data is equivalent to optimizing an upper bound of natural risk on the original data. This implies that adversarial training can serve as a principled defense against delusive attacks. Thus, the test accuracy decreased by delusive attacks can be largely recovered by adversarial training. To further understand the internal mechanism of the defense, we disclose that adversarial training can resist the delusive perturbations by preventing the learner from overly relying on non-robust features in a natural setting. Finally, we complement our theoretical findings with a set of experiments on popular benchmark datasets, which show that the defense withstands six different practical attacks. Both theoretical and empirical results vote for adversarial training when confronted with delusive adversaries.