Unrestricted Adversarial

Unrestricted adversarial attacks aim to create maliciously perturbed data that fool machine learning models while appearing natural to humans, unlike traditional attacks limited by specific perturbation norms. Current research focuses on generating these attacks using diffusion models and generative adversarial networks (GANs), often incorporating techniques like latent space manipulation, semantic guidance from large language models, and recursive token merging for improved realism and transferability across different models. This research is crucial for evaluating the robustness of machine learning systems in real-world scenarios and informing the development of more resilient models and defenses against sophisticated attacks.

Papers

October 3, 2024

SCA: Highly Efficient Semantic-Consistent Unrestricted Adversarial Attack
Zihao Pan, Weibin Wu, Yuhang Cao, Zibin Zheng
Adversarial Example Semantic Adversarial Unrestricted Adversarial

August 10, 2024

ReToMe-VA: Recursive Token Merging for Video Diffusion-based Unrestricted Adversarial Attack
Ziyi Gao, Kai Chen, Zhipeng Wei, Tingshu Mou, Jingjing Chen, Zhiyu Tan, Hao Li, Yu-Gang Jiang
Source Video Token Merging Unrestricted Adversarial Adversarial Consistency Diffusion Attack

April 21, 2024

Reliable Model Watermarking: Defending Against Theft without Compromising on Evasion
Hongyu Zhu, Sichu Liang, Wentao Hu, Fangqi Li, Ju Jia, Shilin Wang
Watermarking Method Adversarial Scenario Model Watermarking Unrestricted Adversarial

April 19, 2024

How Real Is Real? A Human Evaluation Framework for Unrestricted Adversarial Examples
Dren Fazlija, Arkadij Orlov, Johanna Schrader, Monty-Maximilian Zühlke, Michael Rohs, Daniel Kudenko
Adversarial Example Human Evaluation Unrestricted Adversarial Image Attack

April 16, 2024

Efficient Generation of Targeted and Transferable Adversarial Examples for Vision-Language Models Via Diffusion Models
Qi Guo, Shanmin Pang, Xiaojun Jia, Yang Liu, Qing Guo
Diffusion Model Adversarial Example Visual Language Model Adversarial Noise Adversarial Defense Unrestricted Adversarial Transfer Risk

November 27, 2023

RetouchUAA: Unconstrained Adversarial Attack via Image Retouching
Mengda Xie, Yiling He, Meie Fang
Adversarial Attack Adversarial Example Adversarial Image Unrestricted Adversarial Image Retouching Quechua Language

October 6, 2023

Assessing Robustness via Score-Based Adversarial Image Generation
Marcel Kollovieh, Lukas Gosch, Yan Scholten, Marten Lienen, Stephan Günnemann
Adversarial Attack Native Robustness Adversarial Example Score Based Generative Adversarial Generation Unrestricted Adversarial

July 24, 2023

AdvDiff: Generating Unrestricted Adversarial Examples using Diffusion Models
Xuelong Dai, Kaisheng Liang, Bin Xiao
Diffusion Model Adversarial Attack Generative Adversarial Network Adversarial Example Adversarial Defense Unrestricted Adversarial

May 18, 2023

Content-based Unrestricted Adversarial Attack
Zhaoyu Chen, Bo Li, Shuang Wu, Kaixun Jiang, Shouhong Ding, Wenqiang Zhang
Adversarial Attack Adversarial Example Unrestricted Adversarial

January 22, 2023

Provable Unrestricted Adversarial Training without Compromise with Generalizability
Lilin Zhang, Ning Yang, Yanchao Sun, Philip S. Yu
Adversarial Attack Adversarial Training Adversarial Robustness Stronger Generalizability Unrestricted Adversarial Compromise Decision

August 31, 2022

Unrestricted Adversarial Samples Based on Non-semantic Feature Clusters Substitution
MingWei Zhou, Xiaobing Pei
Adversarial Example Adversarial Sample Semantic Feature Defense Method Unrestricted Adversarial Deep Neural Network Classifier

August 24, 2022

Unrestricted Black-box Adversarial Attack Using GAN with Limited Queries
Dongbin Na, Sangwoo Ji, Jong Kim
Adversarial Example GAN Model Unrestricted Adversarial Limited Query

January 4, 2022

Towards Transferable Unrestricted Adversarial Examples with Minimum Changes
Fangcheng Liu, Chao Zhang, Hongyang Zhang
Adversarial Example Adversarial Perturbation Imperceptible Adversarial Perturbation Unrestricted Adversarial Small Change

November 17, 2021

Generating Unrestricted 3D Adversarial Point Clouds
Xuelong Dai, Yanjie Li, Hua Dai, Bin Xiao
Adversarial Point Cloud Unrestricted Adversarial Adversarial Network