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

July 31, 2023

Benchmarking and Analyzing Robust Point Cloud Recognition: Bag of Tricks for Defending Adversarial Examples
Qiufan Ji, Lin Wang, Cong Shi, Shengshan Hu, Yingying Chen, Lichao Sun
Point Cloud Adversarial Example Adversarial Robustness Benchmark Platform Adversarial DEfense Bag Prototype Unconventional Rabbit Hat Trick Point Cloud Recognition Adversarial Robustness Benchmark

July 30, 2023

Theoretically Principled Trade-off for Stateful Defenses against Query-Based Black-Box Attacks
Ashish Hooda, Neal Mangaokar, Ryan Feng, Kassem Fawaz, Somesh Jha, Atul Prakash
Adversarial Example Black Box Attack Optimal Trade Attack Detection Provable Defense

July 26, 2023

July 24, 2023

July 23, 2023

July 21, 2023

July 20, 2023

July 17, 2023

Analyzing the Impact of Adversarial Examples on Explainable Machine Learning
Prathyusha Devabhakthini, Sasmita Parida, Raj Mani Shukla, Suvendu Chandan Nayak
Adversarial Attack Deep Learning Model Adversarial Example Global Impact Adversarial Perturbation Model Interpretability Explainable Machine Learning Text Classification Problem

July 16, 2023

On the Robustness of Split Learning against Adversarial Attacks
Mingyuan Fan, Cen Chen, Chengyu Wang, Wenmeng Zhou, Jun Huang
Adversarial Attack Native Robustness Adversarial Example Split Learning Collaborative Deep Learning Adversarial Color

Adversarial Example

Papers

Benchmarking and Analyzing Robust Point Cloud Recognition: Bag of Tricks for Defending Adversarial Examples

Theoretically Principled Trade-off for Stateful Defenses against Query-Based Black-Box Attacks

R-LPIPS: An Adversarially Robust Perceptual Similarity Metric

NSA: Naturalistic Support Artifact to Boost Network Confidence

FLARE: Fingerprinting Deep Reinforcement Learning Agents using Universal Adversarial Masks

Set-level Guidance Attack: Boosting Adversarial Transferability of Vision-Language Pre-training Models

Enhanced Security against Adversarial Examples Using a Random Ensemble of Encrypted Vision Transformer Models

Lost In Translation: Generating Adversarial Examples Robust to Round-Trip Translation

Gradient-Based Word Substitution for Obstinate Adversarial Examples Generation in Language Models

AdvDiff: Generating Unrestricted Adversarial Examples using Diffusion Models

Robust Automatic Speech Recognition via WavAugment Guided Phoneme Adversarial Training

Towards Generic and Controllable Attacks Against Object Detection

Downstream-agnostic Adversarial Examples

Unveiling Vulnerabilities in Interpretable Deep Learning Systems with Query-Efficient Black-box Attacks

Robust Feature Inference: A Test-time Defense Strategy using Spectral Projections

Improving Transferability of Adversarial Examples via Bayesian Attacks

A LLM Assisted Exploitation of AI-Guardian

Shared Adversarial Unlearning: Backdoor Mitigation by Unlearning Shared Adversarial Examples

Analyzing the Impact of Adversarial Examples on Explainable Machine Learning

On the Robustness of Split Learning against Adversarial Attacks