Test Time Adversarial

Test-time adversarial defense aims to improve the robustness of machine learning models, particularly deep neural networks, against adversarial attacks encountered during deployment without retraining. Current research focuses on developing efficient, training-free methods, such as those leveraging neuron importance ranking or data-free approaches, to enhance model resilience against various attack strategies. These efforts are crucial for ensuring the reliability and safety of AI systems in real-world applications where retraining is impractical or impossible, particularly in safety-critical domains. A key challenge remains achieving robust generalization and mitigating the trade-off between adversarial robustness and standard accuracy.

Papers

October 22, 2024

Test-time Adversarial Defense with Opposite Adversarial Path and High Attack Time Cost
Cheng-Han Yeh, Kuanchun Yu, Chun-Shien Lu
Adversarial Attack Priority User Adversarial Direction Optimal Attack Test Time Adversarial

September 23, 2024

Interpretability-Guided Test-Time Adversarial Defense
Akshay Kulkarni, Tsui-Wei Weng
Neuron Level Test Time Adversarial Test Time Defense

July 25, 2024

Overcoming Binary Adversarial Optimisation with Competitive Coevolution
Per Kristian Lehre, Shishen Lin
Cooperative Coevolution Test Time Adversarial Co Evolutionary Algorithm

June 8, 2024

Bridging the Gap: Rademacher Complexity in Robust and Standard Generalization
Jiancong Xiao, Ruoyu Sun, Qi Long, Weijie J. Su
Adversarial Example Robust Version Robust Loss Rademacher Complexity Test Time Adversarial

January 21, 2024

Efficient local linearity regularization to overcome catastrophic overfitting
Elias Abad Rocamora, Fanghui Liu, Grigorios G. Chrysos, Pablo M. Olmos, Volkan Cevher
Adversarial Perturbation Catastrophic Overfitting Similarity Regularization Single Step Adversarial Training Test Time Adversarial

September 10, 2023

DAD++: Improved Data-free Test Time Adversarial Defense
Gaurav Kumar Nayak, Inder Khatri, Shubham Randive, Ruchit Rawal, Anirban Chakraborty
Adversarial Attack Adversarial Robustness Adversarial DEfense Test Time Adversarial

August 1, 2023

Robust Linear Regression: Phase-Transitions and Precise Tradeoffs for General Norms
Elvis Dohmatob, Meyer Scetbon
Native Robustness Adversarial Robustness Predictive Performance Phase Transition Robust Regression Optimal Trade Learning Norm Fast Minimum Norm Attack Test Time Adversarial

June 26, 2022

Adversarially Robust PAC Learnability of Real-Valued Functions
Idan Attias, Steve Hanneke
Native Robustness Convex Function Vector Valued Fat Shattering Dimension Test Time Adversarial Robust PAC

January 31, 2022

Can Adversarial Training Be Manipulated By Non-Robust Features?
Lue Tao, Lei Feng, Hongxin Wei, Jinfeng Yi, Sheng-Jun Huang, Songcan Chen
Adversarial Training Availability Attack Non Robust Feature Diffusion Attack Test Time Adversarial