Efficient Adversarial Training

Efficient adversarial training aims to enhance the robustness of machine learning models, particularly deep neural networks and large language models (LLMs), against adversarial attacks—maliciously perturbed inputs designed to cause misclassification—without significantly sacrificing accuracy on clean data. Current research focuses on developing faster training algorithms, including those leveraging continuous attacks in embedding spaces, subnetwork sampling, gradient approximation, and data pruning techniques, to reduce computational costs associated with generating adversarial examples. These advancements are crucial for deploying robust models in resource-constrained environments and for improving the reliability and security of AI systems across various applications, from wireless communication to natural language processing.

Papers

August 27, 2024

TART: Boosting Clean Accuracy Through Tangent Direction Guided Adversarial Training
Bongsoo Yi, Rongjie Lai, Yao Li
Adversarial Attack Adversarial Training Multidimensional Local Precision Rate Adversarial DEfense Efficient Adversarial Training Adversarial Direction

May 24, 2024

Efficient Adversarial Training in LLMs with Continuous Attacks
Sophie Xhonneux, Alessandro Sordoni, Stephan Günnemann, Gauthier Gidel, Leo Schwinn
Large Language Model Adversarial Attack Adversarial Training Cyber Attack Fast Adversarial Training Efficient Adversarial Training

April 11, 2024

Adversarial Robustness of Distilled and Pruned Deep Learning-based Wireless Classifiers
Nayan Moni Baishya, B. R. Manoj
Deep Learning Adversarial Attack Adversarial Robustness Better Robustness Signal Classification Automatic Modulation Classification Efficient Adversarial Training

October 24, 2023

Fast Propagation is Better: Accelerating Single-Step Adversarial Training via Sampling Subnetworks
Xiaojun Jia, Jianshu Li, Jindong Gu, Yang Bai, Xiaochun Cao
Adversarial Example Adversarial Training Fast Propagation Step Attack Efficient Adversarial Training Single Step Adversarial Training

September 18, 2023

Reducing Adversarial Training Cost with Gradient Approximation
Huihui Gong
Adversarial Example Adversarial Training Adversarial Loss Gradient Approximation Projected Gradient Descent Efficient Adversarial Training

November 14, 2022

Efficient Adversarial Training with Robust Early-Bird Tickets
Zhiheng Xi, Rui Zheng, Tao Gui, Qi Zhang, Xuanjing Huang
Adversarial Example Pre Trained Language Model Adversarial Training Robust Ticket Efficient Adversarial Training

November 11, 2022

Enhancing and Adversarial: Improve ASR with Speaker Labels
Wei Zhou, Haotian Wu, Jingjing Xu, Mohammad Zeineldeen, Christoph Lüscher, Ralf Schlüter, Hermann Ney
Automatic Speech Recognition Adversarial Training Multi Task Learning Speaker Label Efficient Adversarial Training

October 17, 2022

Probabilistic Categorical Adversarial Attack & Adversarial Training
Han Xu, Pengfei He, Jie Ren, Yuxuan Wan, Zitao Liu, Hui Liu, Jiliang Tang
Adversarial Example Adversarial Training Defense Algorithm Efficient Adversarial Training

October 12, 2022

Efficient Adversarial Training without Attacking: Worst-Case-Aware Robust Reinforcement Learning
Yongyuan Liang, Yanchao Sun, Ruijie Zheng, Furong Huang
Reinforcement Learning Deep Reinforcement Learning Robust Reinforcement Learning Efficient Adversarial Training

October 11, 2022

Stable and Efficient Adversarial Training through Local Linearization
Zhuorong Li, Daiwei Yu
Adversarial Training Iterative Training Catastrophic Overfitting Efficient Adversarial Training Local Linearity Single Step Adversarial Training

July 21, 2022

Towards Efficient Adversarial Training on Vision Transformers
Boxi Wu, Jindong Gu, Zhifeng Li, Deng Cai, Xiaofei He, Wei Liu
Generative Adversarial Network Vision Transformer Fast Adversarial Training Efficient Adversarial Training Robust Convolutional Neural Network

July 1, 2022

Efficient Adversarial Training With Data Pruning
Maximilian Kaufmann, Yiren Zhao, Ilia Shumailov, Robert Mullins, Nicolas Papernot
Adversarial Example Adversarial Training Adversarial Sample Data Pruning Adversarial Evaluation Efficient Adversarial Training

May 22, 2022

AutoJoin: Efficient Adversarial Training against Gradient-Free Perturbations for Robust Maneuvering via Denoising Autoencoder and Joint Learning
Michael Villarreal, Bibek Poudel, Ryan Wickman, Yu Shen, Weizi Li
Adversarial Training Joint Learning Denoising Autoencoder Perception Based Control Efficient Adversarial Training Gradient Perturbation

December 1, 2021

$\ell_\infty$-Robustness and Beyond: Unleashing Efficient Adversarial Training
Hadi M. Dolatabadi, Sarah Erfani, Christopher Leckie
Adversarial Attack Adversarial Example Adversarial Training Fast Adversarial Training Efficient Adversarial Training