Adversarial Example

Adversarial examples are subtly altered inputs designed to fool machine learning models, primarily deep neural networks (DNNs), into making incorrect predictions. Current research focuses on improving model robustness against these attacks, exploring techniques like ensemble methods, multi-objective representation learning, and adversarial training, often applied to architectures such as ResNets and Vision Transformers. Understanding and mitigating the threat of adversarial examples is crucial for ensuring the reliability and security of AI systems across diverse applications, from image classification and natural language processing to malware detection and autonomous driving. The development of robust defenses and effective attack detection methods remains a significant area of ongoing investigation.

Papers

October 19, 2023

ReEval: Automatic Hallucination Evaluation for Retrieval-Augmented Large Language Models via Transferable Adversarial Attacks
Xiaodong Yu, Hao Cheng, Xiaodong Liu, Dan Roth, Jianfeng Gao
Large Language Model Adversarial Example Retrieval Augmentation Transferable Adversarial Attack Hallucination Evaluation

October 18, 2023

October 17, 2023

The Efficacy of Transformer-based Adversarial Attacks in Security Domains
Kunyang Li, Kyle Domico, Jean-Charles Noirot Ferrand, Patrick McDaniel
Adversarial Example Adversarial Perturbation Transformer Architecture Adversarial Sample Adversarial Evaluation Transformer Based Generative Adversarial

October 16, 2023

Black-box Targeted Adversarial Attack on Segment Anything (SAM)
Sheng Zheng, Chaoning Zhang, Xinhong Hao
Adversarial Attack Adversarial Example Segment Anything Model Adversarial Image Segment Anything

October 15, 2023

October 12, 2023

October 11, 2023

October 9, 2023

October 7, 2023

Targeted Attack Improves Protection against Unauthorized Diffusion Customization
Boyang Zheng, Chumeng Liang, Xiaoyu Wu
Adversarial Attack Adversarial Example Latent Diffusion Model

October 6, 2023

October 5, 2023

Adversarial Example

Papers

ReEval: Automatic Hallucination Evaluation for Retrieval-Augmented Large Language Models via Transferable Adversarial Attacks

IRAD: Implicit Representation-driven Image Resampling against Adversarial Attacks

Revisiting Transferable Adversarial Image Examples: Attack Categorization, Evaluation Guidelines, and New Insights

The Efficacy of Transformer-based Adversarial Attacks in Security Domains

Black-box Targeted Adversarial Attack on Segment Anything (SAM)

Towards Deep Learning Models Resistant to Transfer-based Adversarial Attacks via Data-centric Robust Learning

AFLOW: Developing Adversarial Examples under Extremely Noise-limited Settings

SCME: A Self-Contrastive Method for Data-free and Query-Limited Model Extraction Attack

Concealed Electronic Countermeasures of Radar Signal with Adversarial Examples

Effects of Human Adversarial and Affable Samples on BERT Generalization

Boosting Black-box Attack to Deep Neural Networks with Conditional Diffusion Models

An Adversarial Example for Direct Logit Attribution: Memory Management in gelu-4l

An Initial Investigation of Neural Replay Simulator for Over-the-Air Adversarial Perturbations to Automatic Speaker Verification

GReAT: A Graph Regularized Adversarial Training Method

Targeted Attack Improves Protection against Unauthorized Diffusion Customization

Generating Less Certain Adversarial Examples Improves Robust Generalization

Assessing Robustness via Score-Based Adversarial Image Generation

OMG-ATTACK: Self-Supervised On-Manifold Generation of Transferable Evasion Attacks

Adversarial Machine Learning for Social Good: Reframing the Adversary as an Ally

Enhancing Robust Representation in Adversarial Training: Alignment and Exclusion Criteria