Adversarial Transformation

Adversarial transformations explore how small, carefully crafted changes to input data can drastically alter the output of machine learning models, particularly deep neural networks. Current research focuses on developing both attacks (generating these adversarial transformations) and defenses against them, across various data types including images, 3D meshes, and point clouds, employing techniques like spectral decomposition, adversarial training, and generative adversarial networks (GANs). This research is crucial for improving the robustness and reliability of AI systems, with implications for security in applications like face recognition and autonomous driving, as well as for understanding the fundamental limitations of current deep learning architectures.

Papers

October 24, 2024

GADT: Enhancing Transferable Adversarial Attacks through Gradient-guided Adversarial Data Transformation
Yating Ma, Xiaogang Xu, Liming Fang, Zhe Liu
Adversarial Example Adversarial Noise Transferable Attack Transferable Adversarial Attack Adversarial Transformation

January 29, 2023

Towards Verifying the Geometric Robustness of Large-scale Neural Networks
Fu Wang, Peipei Xu, Wenjie Ruan, Xiaowei Huang
Large Scale Verification Task Geometric Transformation Adversarial Transformation

November 24, 2022

SAGA: Spectral Adversarial Geometric Attack on 3D Meshes
Tomer Stolik, Itai Lang, Shai Avidan
Adversarial Attack 3D Mesh Segmentation and Genome Annotation Mesh Autoencoder Spectral Domain Attack Adversarial Transformation Mesh Attack

June 15, 2022

Improving Diversity with Adversarially Learned Transformations for Domain Generalization
Tejas Gokhale, Rushil Anirudh, Jayaraman J. Thiagarajan, Bhavya Kailkhura, Chitta Baral, Yezhou Yang
Domain Generalization Domain Shift Diversity Awareness Learning Transformation Adversarial Transformation Synthetic Source Domain

June 9, 2022

ReFace: Real-time Adversarial Attacks on Face Recognition Systems
Shehzeen Hussain, Todd Huster, Chris Mesterharm, Paarth Neekhara, Kevin An, Malhar Jere, Harshvardhan Sikka, Farinaz Koushanfar
Adversarial Example Adversarial Perturbation Gradient Based Attack G ReFine Adversarial Transformation

April 28, 2022

Improving robustness of language models from a geometry-aware perspective
Bin Zhu, Zhaoquan Gu, Le Wang, Jinyin Chen, Qi Xuan
Language Model Native Robustness Adversarial Training Adversarial Data Enhanced Robustness Adversarial Style Augmentation Adversarial Transformation

February 3, 2022

Learnability Lock: Authorized Learnability Control Through Adversarial Invertible Transformations
Weiqi Peng, Jinghui Chen
Learning Based Attack Adversarial Transformation

November 22, 2021

Imperceptible Transfer Attack and Defense on 3D Point Cloud Classification
Daizong Liu, Wei Hu
Point Cloud Point Cloud Classification Transfer Attack Robust 3D Adversarial Transformation 3D Attack