Physical World Attack

Physical-world attacks target the robustness of machine learning models by introducing real-world perturbations, such as adversarial patches or manipulated lighting, to deceive perception systems like those used in autonomous driving and biometric authentication. Current research focuses on developing increasingly stealthy and effective attacks, exploring various methods including generative adversarial networks (GANs) for anomaly detection and optimization algorithms to design effective yet inconspicuous perturbations, often leveraging natural phenomena like shadows or reflections. These attacks highlight critical vulnerabilities in deployed AI systems, driving the need for improved model robustness and the development of effective countermeasures to ensure safety and security in real-world applications.

Papers

June 9, 2024

April 22, 2024

Swap It Like Its Hot: Segmentation-based spoof attacks on eye-tracking images
Anish S. Narkar, Brendan David-John
Presentation Attack Swap Distance Minimization Liveness Detection Physical World Attack Iris Biometrics

January 7, 2024

Invisible Reflections: Leveraging Infrared Laser Reflections to Target Traffic Sign Perception
Takami Sato, Sri Hrushikesh Varma Bhupathiraju, Michael Clifford, Takeshi Sugawara, Qi Alfred Chen, Sara Rampazzi
Detection Method Traffic Sign Recognition Traffic Sign Man in the Middle Attack Physical World Attack

December 11, 2023

Adversarial Camera Patch: An Effective and Robust Physical-World Attack on Object Detectors
Kalibinuer Tiliwalidi
Object Detector Adversarial Patch Image Patch Physical World Attack Based Physical Attack

November 4, 2023

MTS-DVGAN: Anomaly Detection in Cyber-Physical Systems using a Dual Variational Generative Adversarial Network
Haili Sun, Yan Huang, Lansheng Han, Cai Fu, Hongle Liu, Xiang Long
Generative Model Generative Adversarial Network Anomaly Detection Cyber Physical System Physical World Attack Ct Gan

September 19, 2023

Bridging the Spoof Gap: A Unified Parallel Aggregation Network for Voice Presentation Attacks
Awais Khan, Khalid Mahmood Malik
Aggregation Network Voice Authentication Speech Biomarkers Physical World Attack Voice Cloning Attack

May 22, 2023

Towards Benchmarking and Assessing Visual Naturalness of Physical World Adversarial Attacks
Simin Li, Shuing Zhang, Gujun Chen, Dong Wang, Pu Feng, Jiakai Wang, Aishan Liu, Xin Yi, Xianglong Liu
Adversarial Attack Benchmark Platform Visual Naturalness Physical World Attack Naturalness Assessment Benchmark Attack Enhancement Attack

January 31, 2023

Adversarial Training of Self-supervised Monocular Depth Estimation against Physical-World Attacks
Zhiyuan Cheng, James Liang, Guanhong Tao, Dongfang Liu, Xiangyu Zhang
Adversarial Training Adversarial Robustness Monocular Depth Estimation Self Supervised Monocular Depth Estimation Ground Truth Depth Physical World Attack

December 12, 2022

HOTCOLD Block: Fooling Thermal Infrared Detectors with a Novel Wearable Design
Hui Wei, Zhixiang Wang, Xuemei Jia, Yinqiang Zheng, Hao Tang, Shin'ichi Satoh, Zheng Wang
Adversarial Attack Adversarial Patch Thermal Image Physical World Attack Infrared Detector

September 19, 2022

Adversarial Catoptric Light: An Effective, Stealthy and Robust Physical-World Attack to DNNs
Chengyin Hu, Weiwen Shi
Deep Neural Network Adversarial Perturbation Physical World Attack Based Physical Attack

September 2, 2022

Adversarial Color Film: Effective Physical-World Attack to DNNs
Chengyin Hu, Weiwen Shi
Deep Neural Network Adversarial Training Physical World Attack Adversarial Color Based Physical Attack

June 23, 2022

Adversarial Zoom Lens: A Novel Physical-World Attack to DNNs
Chengyin Hu, Weiwen Shi
Deep Neural Network Adversarial Attack Adversarial Training Physical World Attack

June 2, 2022

Adversarial Laser Spot: Robust and Covert Physical-World Attack to DNNs
Chengyin Hu, Yilong Wang, Kalibinuer Tiliwalidi, Wen Li
Deep Neural Network Physical World Attack Based Physical Attack

March 8, 2022

Shadows can be Dangerous: Stealthy and Effective Physical-world Adversarial Attack by Natural Phenomenon
Yiqi Zhong, Xianming Liu, Deming Zhai, Junjun Jiang, Xiangyang Ji
Adversarial Attack Adversarial Example Rare Event Quantum Shadow Physical World Attack Optical Adversarial

Physical World Attack

Papers

Self-supervised Adversarial Training of Monocular Depth Estimation against Physical-World Attacks

ControlLoc: Physical-World Hijacking Attack on Visual Perception in Autonomous Driving

Swap It Like Its Hot: Segmentation-based spoof attacks on eye-tracking images

Invisible Reflections: Leveraging Infrared Laser Reflections to Target Traffic Sign Perception

Adversarial Camera Patch: An Effective and Robust Physical-World Attack on Object Detectors

MTS-DVGAN: Anomaly Detection in Cyber-Physical Systems using a Dual Variational Generative Adversarial Network

Bridging the Spoof Gap: A Unified Parallel Aggregation Network for Voice Presentation Attacks

Towards Benchmarking and Assessing Visual Naturalness of Physical World Adversarial Attacks

Adversarial Training of Self-supervised Monocular Depth Estimation against Physical-World Attacks

HOTCOLD Block: Fooling Thermal Infrared Detectors with a Novel Wearable Design

Adversarial Catoptric Light: An Effective, Stealthy and Robust Physical-World Attack to DNNs

Adversarial Color Film: Effective Physical-World Attack to DNNs

Adversarial Zoom Lens: A Novel Physical-World Attack to DNNs

Adversarial Laser Spot: Robust and Covert Physical-World Attack to DNNs

Shadows can be Dangerous: Stealthy and Effective Physical-world Adversarial Attack by Natural Phenomenon