Adversarial Training

Adversarial training aims to enhance the robustness of machine learning models, particularly deep neural networks, against adversarial attacks—malicious inputs designed to cause misclassification. Current research focuses on improving the efficiency and effectiveness of adversarial training methods, exploring techniques like vector quantization for input transformation, null-space projection for gradient optimization, and module-wise adaptive training for end-to-end systems, as well as applying these techniques to various model architectures including LLMs and Vision Transformers. This field is crucial for ensuring the reliability and security of AI systems in real-world applications, particularly in safety-critical domains where model robustness is paramount.

Papers

October 10, 2024

Towards Assurance of LLM Adversarial Robustness using Ontology-Driven Argumentation
Tomas Bueno Momcilovic, Beat Buesser, Giulio Zizzo, Mark Purcell, Tomas Bueno Momcilovic
Large Language Model Adversarial Attack Adversarial Training Adversarial Robustness Argumentative Explanation

October 8, 2024

Hyper Adversarial Tuning for Boosting Adversarial Robustness of Pretrained Large Vision Models
Kangtao Lv, Huangsen Cao, Kainan Tu, Yihuai Xu, Zhimeng Zhang, Xin Ding, Yongwei Wang
Adversarial Training Adversarial Robustness Pre Trained Large Vision Model Adversarial Fine Tuning

October 7, 2024

Double Oracle Neural Architecture Search for Game Theoretic Deep Learning Models
Aye Phyu Phyu Aung, Xinrun Wang, Ruiyu Wang, Hau Chan, Bo An, Xiaoli Li, J. Senthilnath
Generative Adversarial Network Adversarial Training GAN Model Game Theoretic

October 4, 2024

Mitigating Adversarial Perturbations for Deep Reinforcement Learning via Vector Quantization
Tung M. Luu, Thanh Nguyen, Tee Joshua Tian Jin, Sungwoon Kim, Chang D. Yoo
Reinforcement Learning Adversarial Attack Deep Reinforcement Learning Adversarial Training Adversarial Perturbation Vector Quantization Robust Agent

October 2, 2024

On Using Certified Training towards Empirical Robustness
Alessandro De Palma, Serge Durand, Zakaria Chihani, François Terrier, Caterina Urban
Adversarial Attack Training Data Adversarial Example Adversarial Training Step Attack Multi Step Attack

September 30, 2024

Robust LLM safeguarding via refusal feature adversarial training
Lei Yu, Virginie Do, Karen Hambardzumyan, Nicola Cancedda
Adversarial Attack Adversarial Training LLM Robustness Refusal Training

September 27, 2024

A3: Active Adversarial Alignment for Source-Free Domain Adaptation
Chrisantus Eze, Christopher Crick
Adversarial Training Unsupervised Domain Adaptation Adversarial Learning Source Free Domain Adaptation Adversarial Alignment

September 26, 2024

Improving Fast Adversarial Training via Self-Knowledge Guidance
Chengze Jiang, Junkai Wang, Minjing Dong, Jie Gui, Xinli Shi, Yuan Cao, Yuan Yan Tang, James Tin-Yau Kwok
Adversarial Attack Adversarial Training Adversarial Robustness Fast Adversarial Training Self Knowledge Efficiency Robustness

September 23, 2024

Improving Adversarial Robustness for 3D Point Cloud Recognition at Test-Time through Purified Self-Training
Jinpeng Lin, Xulei Yang, Tianrui Li, Xun Xu
Adversarial Attack Adversarial Training Adversarial Robustness Self Training Adversarial Sample Test Time Adversarial Purification Point Cloud Recognition

September 19, 2024

Revisiting Semi-supervised Adversarial Robustness via Noise-aware Online Robust Distillation
Tsung-Han Wu, Hung-Ting Su, Shang-Tse Chen, Winston H. Hsu
Training Data Adversarial Training Robust Self Semi Supervised Adversarial

September 18, 2024

NPAT Null-Space Projected Adversarial Training Towards Zero Deterioration
Hanyi Hu, Qiao Han, Kui Chen, Yao Yang
Adversarial Attack Adversarial Training High Performance Null Space Projected Gradient Descent

September 17, 2024

Golden Ratio Search: A Low-Power Adversarial Attack for Deep Learning based Modulation Classification
Deepsayan Sadhukhan, Nitin Priyadarshini Shankar, Sheetal Kalyani
Deep Learning Adversarial Attack Adversarial Training Attack Strategy Modulation Classification Automatic Modulation Classification

September 13, 2024

Exploiting Supervised Poison Vulnerability to Strengthen Self-Supervised Defense
Jeremy Styborski, Mingzhi Lyu, Yi Huang, Adams Kong
Adversarial Training Availability Attack Poisoned Data Poisoning Efficiency

September 11, 2024

Module-wise Adaptive Adversarial Training for End-to-end Autonomous Driving
Tianyuan Zhang, Lu Wang, Jiaqi Kang, Xinwei Zhang, Siyuan Liang, Yuwei Chen, Aishan Liu, Xianglong Liu
Adversarial Attack Autonomous Driving Adversarial Training

September 7, 2024

Phrase-Level Adversarial Training for Mitigating Bias in Neural Network-based Automatic Essay Scoring
Haddad Philip, Tsegaye Misikir Tashu
Adversarial Example Adversarial Training Adversarial Text Mitigating Bias Essay Scoring

August 31, 2024

LightPure: Realtime Adversarial Image Purification for Mobile Devices Using Diffusion Models
Hossein Khalili, Seongbin Park, Vincent Li, Brandan Bright, Ali Payani, Ramana Rao Kompella, Nader Sehatbakhsh
Diffusion Model Adversarial Training Adversarial Robustness Mobile Device Adversarial Purification Adversarial Machine Learning One Shot Generative

August 29, 2024

STEREO: Towards Adversarially Robust Concept Erasing from Text-to-Image Generation Models
Koushik Srivatsan, Fahad Shamshad, Muzammal Naseer, Karthik Nandakumar
Adversarial Attack Adversarial Training Adversarial Prompt Text to Image Generation Model Solar Dynamic Observatory Large Scale Text to Image

August 27, 2024

August 26, 2024

Dual Adversarial Perturbators Generate rich Views for Recommendation
Lijun Zhang, Yuan Yao, Haibo Ye
Contrastive Learning Adversarial Training Adversarial Perturbation Effective Recommendation Best View Contrastive Objective Adversarial Graph

Adversarial Training

Papers

Towards Assurance of LLM Adversarial Robustness using Ontology-Driven Argumentation

Hyper Adversarial Tuning for Boosting Adversarial Robustness of Pretrained Large Vision Models

Double Oracle Neural Architecture Search for Game Theoretic Deep Learning Models

Mitigating Adversarial Perturbations for Deep Reinforcement Learning via Vector Quantization

On Using Certified Training towards Empirical Robustness

Robust LLM safeguarding via refusal feature adversarial training

A3: Active Adversarial Alignment for Source-Free Domain Adaptation

Improving Fast Adversarial Training via Self-Knowledge Guidance

Improving Adversarial Robustness for 3D Point Cloud Recognition at Test-Time through Purified Self-Training

Revisiting Semi-supervised Adversarial Robustness via Noise-aware Online Robust Distillation

NPAT Null-Space Projected Adversarial Training Towards Zero Deterioration

Golden Ratio Search: A Low-Power Adversarial Attack for Deep Learning based Modulation Classification

Exploiting Supervised Poison Vulnerability to Strengthen Self-Supervised Defense

Module-wise Adaptive Adversarial Training for End-to-end Autonomous Driving

Phrase-Level Adversarial Training for Mitigating Bias in Neural Network-based Automatic Essay Scoring

LightPure: Realtime Adversarial Image Purification for Mobile Devices Using Diffusion Models

STEREO: Towards Adversarially Robust Concept Erasing from Text-to-Image Generation Models

Adversarial Attacks and Defenses in Multivariate Time-Series Forecasting for Smart and Connected Infrastructures

TART: Boosting Clean Accuracy Through Tangent Direction Guided Adversarial Training

Dual Adversarial Perturbators Generate rich Views for Recommendation