# Bilateral Adversarial Training: Towards Fast Training of More Robust Models Against Adversarial Attacks

arXiv: https://arxiv.org/abs/1811.10716 · HTML: https://mlsanity.com/paper/1811.10716/

## Key facts

- Cited by 48 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 26 Nov 2018.

## Built on by (most cited first, 30 of 48)

- [Recent Advances in Adversarial Training for Adversarial Robustness](https://mlsanity.com/paper/2102.01356/) (2021)
- [Advances in adversarial attacks and defenses in computer vision: A survey](https://mlsanity.com/paper/2108.00401/) (2021)
- [SegPGD: An Effective and Efficient Adversarial Attack for Evaluating and Boosting Segmentation Robustness](https://mlsanity.com/paper/2207.12391/) (2022)
- [How Deep Learning Sees the World: A Survey on Adversarial Attacks & Defenses](https://mlsanity.com/paper/2305.10862/) (2023)
- [A Survey of Robust Adversarial Training in Pattern Recognition: Fundamental, Theory, and Methodologies](https://mlsanity.com/paper/2203.14046/) (2022)
- [Sparsity Winning Twice: Better Robust Generalization from More Efficient Training](https://mlsanity.com/paper/2202.09844/) (2022)
- [DISCO: Adversarial Defense with Local Implicit Functions](https://mlsanity.com/paper/2212.05630/) (2022)
- [Randomized Adversarial Training via Taylor Expansion](https://mlsanity.com/paper/2303.10653/) (2023)
- [Investigating Top-k White-Box and Transferable Black-box Attack](https://mlsanity.com/paper/2204.00089/) (2022)
- [Understanding and Combating Robust Overfitting via Input Loss Landscape Analysis and Regularization](https://mlsanity.com/paper/2212.04985/) (2022)
- [Improving the Accuracy-Robustness Trade-Off of Classifiers via Adaptive Smoothing](https://mlsanity.com/paper/2301.12554/) (2023)
- [Fast Adversarial Training with Smooth Convergence](https://mlsanity.com/paper/2308.12857/) (2023)
- [Fooling Adversarial Training with Inducing Noise](https://mlsanity.com/paper/2111.10130/) (2021)
- [AGKD-BML: Defense Against Adversarial Attack by Attention Guided Knowledge Distillation and Bi-directional Metric Learning](https://mlsanity.com/paper/2108.06017/) (2021)
- [Enhancing Adversarial Robustness for Deep Metric Learning](https://mlsanity.com/paper/2203.01439/) (2022)
- [Improving Model Robustness with Latent Distribution Locally and Globally](https://mlsanity.com/paper/2107.04401/) (2021)
- [Adversarial Attack and Defense in Deep Ranking](https://mlsanity.com/paper/2106.03614/) (2021)
- [Real World Federated Learning with a Knowledge Distilled Transformer for Cardiac CT Imaging](https://mlsanity.com/paper/2407.07557/) (2024)
- [Adversarial Finetuning with Latent Representation Constraint to Mitigate Accuracy-Robustness Tradeoff](https://mlsanity.com/paper/2308.16454/) (2023)
- [Provable Unrestricted Adversarial Training without Compromise with Generalizability](https://mlsanity.com/paper/2301.09069/) (2023)
- [Adversarial Training: A Survey](https://mlsanity.com/paper/2410.15042/) (2024)
- [All You Need is RAW: Defending Against Adversarial Attacks with Camera Image Pipelines](https://mlsanity.com/paper/2112.09219/) (2021)
- [Adversarial Attacks and Defenses in Machine Learning-Powered Networks: A Contemporary Survey](https://mlsanity.com/paper/2303.06302/) (2023)
- [Towards Robust and Accurate Visual Prompting](https://mlsanity.com/paper/2311.10992/) (2023)
- [Towards Speeding up Adversarial Training in Latent Spaces](https://mlsanity.com/paper/2102.00662/) (2021)
- [Adversarial Attacks and Defenses for Social Network Text Processing Applications: Techniques, Challenges and Future Research Directions](https://mlsanity.com/paper/2110.13980/) (2021)
- [A3T: Accuracy Aware Adversarial Training](https://mlsanity.com/paper/2211.16316/) (2022)
- [Exploring Robust Features for Improving Adversarial Robustness](https://mlsanity.com/paper/2309.04650/) (2023)
- [Adaptive Feature Alignment for Adversarial Training](https://mlsanity.com/paper/2105.15157/) (2021)
- [Adversarial Attacks on ML Defense Models Competition](https://mlsanity.com/paper/2110.08042/) (2021)

## Abstract (arXiv)

In this paper, we study fast training of adversarially robust models. From the analyses of the state-of-the-art defense method, i.e., the multi-step adversarial training, we hypothesize that the gradient magnitude links to the model robustness. Motivated by this, we propose to perturb both the image and the label during training, which we call Bilateral Adversarial Training (BAT). To generate the adversarial label, we derive an closed-form heuristic solution. To generate the adversarial image, we use one-step targeted attack with the target label being the most confusing class. In the experiment, we first show that random start and the most confusing target attack effectively prevent the label leaking and gradient masking problem. Then coupled with the adversarial label part, our model significantly improves the state-of-the-art results. For example, against PGD100 white-box attack with cross-entropy loss, on CIFAR10, we achieve 63.7% versus 47.2%; on SVHN, we achieve 59.1% versus 42.1%. At last, the experiment on the very (computationally) challenging ImageNet dataset further demonstrates the effectiveness of our fast method.