Explaining and Harnessing Adversarial Examples

Abstract

Several machine learning models, including neural networks, consistently misclassify adversarial examples---inputs formed by applying small but intentionally worst-case perturbations to examples from the dataset, such that the perturbed input results in the model outputting an incorrect answer with high confidence. Early attempts at explaining this phenomenon focused on nonlinearity and overfitting. We argue instead that the primary cause of neural networks' vulnerability to adversarial perturbation is their linear nature. This explanation is supported by new quantitative results while giving the first explanation of the most intriguing fact about them: their generalization across architectures and training sets. Moreover, this view yields a simple and fast method of generating adversarial examples. Using this approach to provide examples for adversarial training, we reduce the test set error of a maxout network on the MNIST dataset.

References in corpus (5)

• Intriguing properties of neural networks
• Going Deeper with Convolutions
• Theano: new features and speed improvements
• Towards Deep Neural Network Architectures Robust to Adversarial Examples
• Pylearn2: a machine learning research library

Cited by in corpus (814)

• Feature Squeezing: Detecting Adversarial Examples in Deep Neural Networks
• ZOO: Zeroth Order Optimization based Black-box Attacks to Deep Neural Networks without Training Substitute Models
• Explaining Deep Neural Networks and Beyond: A Review of Methods and Applications
• NIPS 2016 Tutorial: Generative Adversarial Networks
• A trans-disciplinary review of deep learning research for water resources scientists
• Recent Advances in Open Set Recognition: A Survey
• BadNets: Identifying Vulnerabilities in the Machine Learning Model Supply Chain
• Robust Real-World Image Super-Resolution against Adversarial Attacks
• Explaining Explanations in AI
• Blended Diffusion for Text-driven Editing of Natural Images
• On Evaluating Adversarial Robustness
• The Hidden Vulnerability of Distributed Learning in Byzantium
• Understanding Adversarial Attacks on Deep Learning Based Medical Image Analysis Systems
• Adversarially Robust Generalization Requires More Data
• MedViT: A Robust Vision Transformer for Generalized Medical Image Classification
• Data Augmentation using Random Image Cropping and Patching for Deep CNNs
• Smart Augmentation - Learning an Optimal Data Augmentation Strategy
• The Challenge of Machine Learning in Space Weather Nowcasting and Forecasting
• Adversarial Personalized Ranking for Recommendation
• Deep Text Classification Can be Fooled
• Understanding Adversarial Training: Increasing Local Stability of Neural Nets through Robust Optimization
• Vertical Federated Learning: Concepts, Advances and Challenges
• Towards Principled Methods for Training Generative Adversarial Networks
• Deep and Confident Prediction for Time Series at Uber
• A Closer Look at Memorization in Deep Networks
• Image Classification with Deep Learning in the Presence of Noisy Labels: A Survey
• Adversarial Frontier Stitching for Remote Neural Network Watermarking
• Unleashing the potential of prompt engineering for large language models
• TextBugger: Generating Adversarial Text Against Real-world Applications
• Generalized Out-of-Distribution Detection: A Survey
• Adversarial Attack and Defense on Graph Data: A Survey
• A Survey of End-to-End Driving: Architectures and Training Methods
• AdvHat: Real-world adversarial attack on ArcFace Face ID system
• Uncertainty-aware multi-view co-training for semi-supervised medical image segmentation and domain adaptation
• Compact Convolutional Neural Networks for Classification of Asynchronous Steady-state Visual Evoked Potentials
• Pathologies of Neural Models Make Interpretations Difficult
• X-ModalNet: A Semi-Supervised Deep Cross-Modal Network for Classification of Remote Sensing Data
• A study of the effect of JPG compression on adversarial images
• Februus: Input Purification Defense Against Trojan Attacks on Deep Neural Network Systems
• Var-CNN: A Data-Efficient Website Fingerprinting Attack Based on Deep Learning
• Neural Trojans
• The Fallacy of AI Functionality
• Adversarial Examples: Opportunities and Challenges
• GasHis-Transformer: A Multi-scale Visual Transformer Approach for Gastric Histopathological Image Detection
• Keeping the Bad Guys Out: Protecting and Vaccinating Deep Learning with JPEG Compression
• Invisible for both Camera and LiDAR: Security of Multi-Sensor Fusion based Perception in Autonomous Driving Under Physical-World Attacks
• Unveiling the potential of Graph Neural Networks for network modeling and optimization in SDN
• Multiplicative Normalizing Flows for Variational Bayesian Neural Networks
• Hardware and Software Optimizations for Accelerating Deep Neural Networks: Survey of Current Trends, Challenges, and the Road Ahead
• A Comprehensive Survey of Deep Transfer Learning for Anomaly Detection in Industrial Time Series: Methods, Applications, and Directions
• Security and Privacy Approaches in Mixed Reality: A Literature Survey
• Qualitatively characterizing neural network optimization problems
• Transferring Robustness for Graph Neural Network Against Poisoning Attacks
• Evaluating and Improving Adversarial Robustness of Machine Learning-Based Network Intrusion Detectors
• ShakeDrop Regularization for Deep Residual Learning
• Learning to Compose Domain-Specific Transformations for Data Augmentation
• Adversarial Example Detection for DNN Models: A Review and Experimental Comparison
• Direct Feedback Alignment Provides Learning in Deep Neural Networks
• Fast Yet Effective Machine Unlearning
• Spectrally-normalized margin bounds for neural networks
• Are Labels Required for Improving Adversarial Robustness?
• Robust Audio Adversarial Example for a Physical Attack
• Ten Years of Generative Adversarial Nets (GANs): A survey of the state-of-the-art
• Towards Crafting Text Adversarial Samples
• Natural Attack for Pre-trained Models of Code
• Quantum noise protects quantum classifiers against adversaries
• Evaluating the Robustness of Neural Networks: An Extreme Value Theory Approach
• Feature relevance quantification in explainable AI: A causal problem
• Copycat CNN: Stealing Knowledge by Persuading Confession with Random Non-Labeled Data
• SpaText: Spatio-Textual Representation for Controllable Image Generation
• AdvPC: Transferable Adversarial Perturbations on 3D Point Clouds
• The Fishyscapes Benchmark: Measuring Blind Spots in Semantic Segmentation
• Adversarial Attacks on Deep Neural Networks for Time Series Classification
• FETCH: A deep-learning based classifier for fast transient classification
• advertorch v0.1: An Adversarial Robustness Toolbox based on PyTorch
• PrivacyNet: Semi-Adversarial Networks for Multi-attribute Face Privacy
• Ensemble Kalman Inversion: A Derivative-Free Technique For Machine Learning Tasks
• Training robust neural networks using Lipschitz bounds
• How Deep Learning Sees the World: A Survey on Adversarial Attacks & Defenses
• A Boundary Tilting Persepective on the Phenomenon of Adversarial Examples
• Adversarial EXEmples: A Survey and Experimental Evaluation of Practical Attacks on Machine Learning for Windows Malware Detection
• Adversarial Attack Vulnerability of Medical Image Analysis Systems: Unexplored Factors
• Accountability in an Algorithmic Society: Relationality, Responsibility, and Robustness in Machine Learning
• Non-convex Min-Max Optimization: Applications, Challenges, and Recent Theoretical Advances
• Adversarial Examples, Uncertainty, and Transfer Testing Robustness in Gaussian Process Hybrid Deep Networks
• Learning to Protect Communications with Adversarial Neural Cryptography
• Decision-Focused Learning: Foundations, State of the Art, Benchmark and Future Opportunities
• Fooling Neural Network Interpretations via Adversarial Model Manipulation
• AI Safety Gridworlds
• Supervised Contrastive Learning
• Controversial stimuli: pitting neural networks against each other as models of human recognition
• A Survey on AI Assurance
• Manipulating Machine Learning: Poisoning Attacks and Countermeasures for Regression Learning
• Guessing Smart: Biased Sampling for Efficient Black-Box Adversarial Attacks
• Self-Adversarial Training incorporating Forgery Attention for Image Forgery Localization
• Dropout Inference in Bayesian Neural Networks with Alpha-divergences
• Interpreting and Improving Adversarial Robustness of Deep Neural Networks with Neuron Sensitivity
• Review: Deep Learning in Electron Microscopy
• Are All Layers Created Equal?
• Fast Feature Fool: A data independent approach to universal adversarial perturbations
• Verification of Neural Network Behaviour: Formal Guarantees for Power System Applications
• Balanced Semi-Supervised Generative Adversarial Network for Damage Assessment from Low-Data Imbalanced-Class Regime
• TSViz: Demystification of Deep Learning Models for Time-Series Analysis
• Benchmarking Adversarial Patch Against Aerial Detection
• Universal Adversarial Examples in Remote Sensing: Methodology and Benchmark
• Label-Only Membership Inference Attacks
• LESSON: Multi-Label Adversarial False Data Injection Attack for Deep Learning Locational Detection
• Toward Open-World Electroencephalogram Decoding Via Deep Learning: A Comprehensive Survey
• Understanding and Enhancing the Transferability of Adversarial Examples
• Adversarial Attacks Against Deep Generative Models on Data: A Survey
• How to Certify Machine Learning Based Safety-critical Systems? A Systematic Literature Review
• Analyzing the Robustness of Nearest Neighbors to Adversarial Examples
• It's Morphin' Time! Combating Linguistic Discrimination with Inflectional Perturbations
• Practical Blind Membership Inference Attack via Differential Comparisons
• Denoising diffusion algorithm for inverse design of microstructures with fine-tuned nonlinear material properties
• Universal adversarial perturbations
• Traffic Sign Detection under Challenging Conditions: A Deeper Look Into Performance Variations and Spectral Characteristics
• Towards Interpretable Deep Neural Networks by Leveraging Adversarial Examples
• Physics-Driven Regularization of Deep Neural Networks for Enhanced Engineering Design and Analysis
• Improving the Robustness of Deep Neural Networks via Stability Training
• Assuring the Machine Learning Lifecycle: Desiderata, Methods, and Challenges
• Attacking the Madry Defense Model with $L_1$-based Adversarial Examples
• Secure and Trustworthy Artificial Intelligence-Extended Reality (AI-XR) for Metaverses
• Optimization and Abstraction: A Synergistic Approach for Analyzing Neural Network Robustness
• Improving the Generalization of Adversarial Training with Domain Adaptation
• Towards Adversarial Malware Detection: Lessons Learned from PDF-based Attacks
• Robust Android Malware Detection System against Adversarial Attacks using Q-Learning
• How is Gaze Influenced by Image Transformations? Dataset and Model
• AI Security for Geoscience and Remote Sensing: Challenges and Future Trends
• Photographic Image Synthesis with Cascaded Refinement Networks
• Towards quantum enhanced adversarial robustness in machine learning
• Graph Neural Networks with Continual Learning for Fake News Detection from Social Media
• Explaining Vulnerabilities to Adversarial Machine Learning through Visual Analytics
• Scattering Networks for Hybrid Representation Learning
• Launching Adversarial Attacks against Network Intrusion Detection Systems for IoT
• Does Your Dermatology Classifier Know What It Doesn't Know? Detecting the Long-Tail of Unseen Conditions
• Semantic Robustness of Models of Source Code
• Intellectual Property Protection for Deep Learning Models: Taxonomy, Methods, Attacks, and Evaluations
• Backdoor Pre-trained Models Can Transfer to All
• Gotta Catch 'Em All: Using Honeypots to Catch Adversarial Attacks on Neural Networks
• Revisiting Adversarially Learned Injection Attacks Against Recommender Systems
• Adversarial Perturbations Against Real-Time Video Classification Systems
• A Counter-Forensic Method for CNN-Based Camera Model Identification
• OpenAttack: An Open-source Textual Adversarial Attack Toolkit
• Revisiting LSTM Networks for Semi-Supervised Text Classification via Mixed Objective Function
• Non-Intrusive Reduced-Order Modeling Using Uncertainty-Aware Deep Neural Networks and Proper Orthogonal Decomposition: Application to Flood Modeling
• Facial Attributes: Accuracy and Adversarial Robustness
• A Simple Explanation for the Existence of Adversarial Examples with Small Hamming Distance
• Malware Makeover: Breaking ML-based Static Analysis by Modifying Executable Bytes
• A newcomer's guide to deep learning for inverse design in nano-photonics
• Seq2Sick: Evaluating the Robustness of Sequence-to-Sequence Models with Adversarial Examples
• Synesthesia: Detecting Screen Content via Remote Acoustic Side Channels
• A Survey on Reservoir Computing and its Interdisciplinary Applications Beyond Traditional Machine Learning
• Adversarial Manipulation of Deep Representations
• Promises and pitfalls of deep neural networks in neuroimaging-based psychiatric research
• FDA3 : Federated Defense Against Adversarial Attacks for Cloud-Based IIoT Applications
• 3D-VField: Adversarial Augmentation of Point Clouds for Domain Generalization in 3D Object Detection
• When Does Self-Supervision Help Graph Convolutional Networks?
• Privacy Intelligence: A Survey on Image Privacy in Online Social Networks
• Robust Federated Learning: The Case of Affine Distribution Shifts
• Partial success in closing the gap between human and machine vision
• EAD: Elastic-Net Attacks to Deep Neural Networks via Adversarial Examples
• Characterizing Speech Adversarial Examples Using Self-Attention U-Net Enhancement
• Decay Replay Mining to Predict Next Process Events
• Seabed classification using physics-based modeling and machine learning
• Improving the Transferability of Adversarial Examples with Resized-Diverse-Inputs, Diversity-Ensemble and Region Fitting
• Can Adversarial Weight Perturbations Inject Neural Backdoors?
• Maximal Jacobian-based Saliency Map Attack
• The Limitations of Adversarial Training and the Blind-Spot Attack
• Vulnerability of deep neural networks for detecting COVID-19 cases from chest X-ray images to universal adversarial attacks
• On Lightweight Privacy-Preserving Collaborative Learning for IoT Objects
• Towards Evaluating the Robustness of Deep Diagnostic Models by Adversarial Attack
• Machine Learning in NextG Networks via Generative Adversarial Networks
• Making DeepFakes more spurious: evading deep face forgery detection via trace removal attack
• Characterizing and evaluating adversarial examples for Offline Handwritten Signature Verification
• Watch out! Motion is Blurring the Vision of Your Deep Neural Networks
• Explanation-Guided Backdoor Poisoning Attacks Against Malware Classifiers
• Detection of Face Recognition Adversarial Attacks
• Are Accuracy and Robustness Correlated?
• Unsupervised Histopathology Image Synthesis
• Discretization based Solutions for Secure Machine Learning against Adversarial Attacks
• A Survey of Safety and Trustworthiness of Deep Neural Networks: Verification, Testing, Adversarial Attack and Defence, and Interpretability
• Adversarial Robustness of Deep Neural Networks: A Survey from a Formal Verification Perspective
• Recent Advances of Differential Privacy in Centralized Deep Learning: A Systematic Survey
• AdvMIL: Adversarial Multiple Instance Learning for the Survival Analysis on Whole-Slide Images
• An Alternative Surrogate Loss for PGD-based Adversarial Testing
• Training individually fair ML models with Sensitive Subspace Robustness
• RobustART: Benchmarking Robustness on Architecture Design and Training Techniques
• Adversarial Dropout Regularization
• Detecting Adversarial Examples by Input Transformations, Defense Perturbations, and Voting
• WAF-A-MoLE: Evading Web Application Firewalls through Adversarial Machine Learning
• Boosting Adversarial Attacks with Momentum
• Concise Fuzzy System Modeling Integrating Soft Subspace Clustering and Sparse Learning
• Inspect, Understand, Overcome: A Survey of Practical Methods for AI Safety
• ModelDiff: Testing-Based DNN Similarity Comparison for Model Reuse Detection
• Noise Flooding for Detecting Audio Adversarial Examples Against Automatic Speech Recognition
• Yes, we GAN: Applying Adversarial Techniques for Autonomous Driving
• CBA: Contextual Background Attack against Optical Aerial Detection in the Physical World
• FedDef: Defense Against Gradient Leakage in Federated Learning-based Network Intrusion Detection Systems
• Robust in Practice: Adversarial Attacks on Quantum Machine Learning
• Shield: Fast, Practical Defense and Vaccination for Deep Learning using JPEG Compression
• VisualBackProp: efficient visualization of CNNs
• Threatening Patch Attacks on Object Detection in Optical Remote Sensing Images
• I-MAD: Interpretable Malware Detector Using Galaxy Transformer
• Fault Sneaking Attack: a Stealthy Framework for Misleading Deep Neural Networks
• Efficient and Accurate Estimation of Lipschitz Constants for Deep Neural Networks
• Sparse Unsupervised Capsules Generalize Better
• Understanding trained CNNs by indexing neuron selectivity
• Towards Robust Neural Image Compression: Adversarial Attack and Model Finetuning
• Entangled Watermarks as a Defense against Model Extraction
• Defending Against Universal Attacks Through Selective Feature Regeneration
• Evading Defenses to Transferable Adversarial Examples by Translation-Invariant Attacks
• Adversarially Regularising Neural NLI Models to Integrate Logical Background Knowledge
• Backdoors in Neural Models of Source Code
• Adaptative Perturbation Patterns: Realistic Adversarial Learning for Robust Intrusion Detection
• Rigorous Agent Evaluation: An Adversarial Approach to Uncover Catastrophic Failures
• VC Classes are Adversarially Robustly Learnable, but Only Improperly
• Semantic-Aware Adversarial Training for Reliable Deep Hashing Retrieval
• A deep architecture for unified aesthetic prediction
• Technical report on Conversational Question Answering
• Deep Weighted Averaging Classifiers
• Weighted-Sampling Audio Adversarial Example Attack
• Advancing diagnostic performance and clinical usability of neural networks via adversarial training and dual batch normalization
• Adversarial Patch Attacks on Monocular Depth Estimation Networks
• Interpretable neural networks based on continuous-valued logic and multicriteria decision operators
• Indirect Adversarial Attacks via Poisoning Neighbors for Graph Convolutional Networks
• On Adversarial Examples and Stealth Attacks in Artificial Intelligence Systems
• DeepBillboard: Systematic Physical-World Testing of Autonomous Driving Systems
• Adversarial Machine Learning And Speech Emotion Recognition: Utilizing Generative Adversarial Networks For Robustness
• SemMT: A Semantic-based Testing Approach for Machine Translation Systems
• Advbox: a toolbox to generate adversarial examples that fool neural networks
• Red Alarm for Pre-trained Models: Universal Vulnerability to Neuron-Level Backdoor Attacks
• Robustness Verification of Quantum Classifiers
• The Vulnerability of Semantic Segmentation Networks to Adversarial Attacks in Autonomous Driving: Enhancing Extensive Environment Sensing
• Just Noticeable Difference for Deep Machine Vision
• Reward learning from human preferences and demonstrations in Atari
• Stealing Links from Graph Neural Networks
• Biometric Backdoors: A Poisoning Attack Against Unsupervised Template Updating
• On the Geometry of Adversarial Examples
• Understanding Neural Code Intelligence Through Program Simplification
• Adversarial classification: An adversarial risk analysis approach
• Adversarial Driving: Attacking End-to-End Autonomous Driving
• Local minima in training of neural networks
• Machine Learning (In) Security: A Stream of Problems
• Adversarially Robust Neural Architectures
• RoMFAC: A robust mean-field actor-critic reinforcement learning against adversarial perturbations on states
• Multi-Agent Reinforcement Learning: Methods, Applications, Visionary Prospects, and Challenges
• The worst of both worlds: A comparative analysis of errors in learning from data in psychology and machine learning
• On the Robustness of Convolutional Neural Networks to Internal Architecture and Weight Perturbations
• Get your Foes Fooled: Proximal Gradient Split Learning for Defense against Model Inversion Attacks on IoMT data
• Generation & Evaluation of Adversarial Examples for Malware Obfuscation
• AdvSQLi: Generating Adversarial SQL Injections against Real-world WAF-as-a-service
• Spinning Language Models: Risks of Propaganda-As-A-Service and Countermeasures
• Humans and deep networks largely agree on which kinds of variation make object recognition harder
• PatchUp: A Feature-Space Block-Level Regularization Technique for Convolutional Neural Networks
• Exploiting Cultural Biases via Homoglyphs in Text-to-Image Synthesis
• Defensive Quantization: When Efficiency Meets Robustness
• Calibration of Deep Probabilistic Models with Decoupled Bayesian Neural Networks
• PAD: Towards Principled Adversarial Malware Detection Against Evasion Attacks
• Large batch size training of neural networks with adversarial training and second-order information
• Mimic and Fool: A Task Agnostic Adversarial Attack
• Quality Resilient Deep Neural Networks
• Adversarial Examples Make Strong Poisons
• Tensor Graph Convolutional Networks for Multi-relational and Robust Learning
• Pixle: a fast and effective black-box attack based on rearranging pixels
• Markov Chain Monte Carlo-Based Machine Unlearning: Unlearning What Needs to be Forgotten
• Understanding Real-world Threats to Deep Learning Models in Android Apps
• Interpreting Black Box Models via Hypothesis Testing
• Toward Robust Image Classification
• Bypassing Feature Squeezing by Increasing Adversary Strength
• Adversarial Example in Remote Sensing Image Recognition
• Quantifying Perceptual Distortion of Adversarial Examples
• On the Adversarial Robustness of LASSO Based Feature Selection
• White Box Methods for Explanations of Convolutional Neural Networks in Image Classification Tasks
• Towards Universal Physical Attacks On Cascaded Camera-Lidar 3D Object Detection Models
• Towards a Robust and Trustworthy Machine Learning System Development: An Engineering Perspective
• Information Aware Max-Norm Dirichlet Networks for Predictive Uncertainty Estimation
• Improving the Transferability of Adversarial Examples with Arbitrary Style Transfer
• Bad Global Minima Exist and SGD Can Reach Them
• Mitigating Adversarial Gray-Box Attacks Against Phishing Detectors
• Greedy Attack and Gumbel Attack: Generating Adversarial Examples for Discrete Data
• Attention Meets Perturbations: Robust and Interpretable Attention with Adversarial Training
• Extracting Sentence Embeddings from Pretrained Transformer Models
• On the Limitation of Local Intrinsic Dimensionality for Characterizing the Subspaces of Adversarial Examples
• Adversarial Multi-task Learning Enhanced Physics-informed Neural Networks for Solving Partial Differential Equations
• Software Testing for Machine Learning
• Enabling AutoML for Zero-Touch Network Security: Use-Case Driven Analysis
• Machine Learning with Confidential Computing: A Systematization of Knowledge
• When and How to Fool Explainable Models (and Humans) with Adversarial Examples
• The Future of Misinformation Detection: New Perspectives and Trends
• Adversarial Robustness via Fisher-Rao Regularization
• Interpretability Beyond Classification Output: Semantic Bottleneck Networks
• Training-free Lexical Backdoor Attacks on Language Models
• Increasing the Confidence of Deep Neural Networks by Coverage Analysis
• "That Is a Suspicious Reaction!": Interpreting Logits Variation to Detect NLP Adversarial Attacks
• PEFT-SER: On the Use of Parameter Efficient Transfer Learning Approaches For Speech Emotion Recognition Using Pre-trained Speech Models
• Adversarial Phenomenon in the Eyes of Bayesian Deep Learning
• Towards out of distribution generalization for problems in mechanics
• AI Research Considerations for Human Existential Safety (ARCHES)
• Interpolated Adversarial Training: Achieving Robust Neural Networks without Sacrificing Too Much Accuracy
• Towards Resilient Artificial Intelligence: Survey and Research Issues
• Adversarial Examples in Deep Learning: Characterization and Divergence
• Task dependent Deep LDA pruning of neural networks
• MPSN: Motion-aware Pseudo Siamese Network for Indoor Video Head Detection in Buildings
• Sensitivity analysis of Wasserstein distributionally robust optimization problems
• Investigating the significance of adversarial attacks and their relation to interpretability for radar-based human activity recognition systems
• Towards Frequency-Based Explanation for Robust CNN
• TrISec: Training Data-Unaware Imperceptible Security Attacks on Deep Neural Networks
• Improving Neural Relation Extraction with Implicit Mutual Relations
• NetFense: Adversarial Defenses against Privacy Attacks on Neural Networks for Graph Data
• Predify: Augmenting deep neural networks with brain-inspired predictive coding dynamics
• Two Sides of the Same Coin: White-box and Black-box Attacks for Transfer Learning
• The RFML Ecosystem: A Look at the Unique Challenges of Applying Deep Learning to Radio Frequency Applications
• Gerrymandering and Compactness: Implementation Flexibility and Abuse
• Explaining Vulnerabilities of Deep Learning to Adversarial Malware Binaries
• NeurObfuscator: A Full-stack Obfuscation Tool to Mitigate Neural Architecture Stealing
• Mutual Adversarial Training: Learning together is better than going alone
• ECGadv: Generating Adversarial Electrocardiogram to Misguide Arrhythmia Classification System
• Attack Analysis of Face Recognition Authentication Systems Using Fast Gradient Sign Method
• Probabilistic Deep Learning to Quantify Uncertainty in Air Quality Forecasting
• Regularization Learning Networks: Deep Learning for Tabular Datasets
• Adversarial Robustness of Neural-Statistical Features in Detection of Generative Transformers
• Global Robustness Evaluation of Deep Neural Networks with Provable Guarantees for the $L_0$ Norm
• Morphence: Moving Target Defense Against Adversarial Examples
• Attacks on State-of-the-Art Face Recognition using Attentional Adversarial Attack Generative Network
• SemiCurv: Semi-Supervised Curvilinear Structure Segmentation
• Defending Person Detection Against Adversarial Patch Attack by using Universal Defensive Frame
• High Dimensional Spaces, Deep Learning and Adversarial Examples
• Adversarial Machine Learning: Bayesian Perspectives
• Radio Frequency Fingerprinting via Deep Learning: Challenges and Opportunities
• Improving the Adversarial Robustness of NLP Models by Information Bottleneck
• Towards Visual Distortion in Black-Box Attacks
• Constrained Learning with Non-Convex Losses
• DANCin SEQ2SEQ: Fooling Text Classifiers with Adversarial Text Example Generation
• Data-driven Identification of 2D Partial Differential Equations using extracted physical features
• Defend Data Poisoning Attacks on Voice Authentication
• Black-box adversarial attacks using Evolution Strategies
• TMS-Net: A Segmentation Network Coupled With A Run-time Quality Control Method For Robust Cardiac Image Segmentation
• Fooling a Real Car with Adversarial Traffic Signs
• Action-Manipulation Attacks Against Stochastic Bandits: Attacks and Defense
• When NAS Meets Robustness: In Search of Robust Architectures against Adversarial Attacks
• Towards Imperceptible and Robust Adversarial Example Attacks against Neural Networks
• Adversarial regression training for visualizing the progression of chronic obstructive pulmonary disease with chest x-rays
• The Faults in Our Pi Stars: Security Issues and Open Challenges in Deep Reinforcement Learning
• Formal Verification of CNN-based Perception Systems
• Backdoor Attacks for Remote Sensing Data with Wavelet Transform
• Prompt Perturbation in Retrieval-Augmented Generation based Large Language Models
• Improving Back-Propagation by Adding an Adversarial Gradient
• Provable Certificates for Adversarial Examples: Fitting a Ball in the Union of Polytopes
• Simple black-box universal adversarial attacks on medical image classification based on deep neural networks
• What You See is Not What the Network Infers: Detecting Adversarial Examples Based on Semantic Contradiction
• Evaluating the visualization of what a Deep Neural Network has learned
• Beneficial Perturbation Network for designing general adaptive artificial intelligence systems
• Safety Verification of Neural Feedback Systems Based on Constrained Zonotopes
• GAN-Aimbots: Using Machine Learning for Cheating in First Person Shooters
• Rethinking Non-idealities in Memristive Crossbars for Adversarial Robustness in Neural Networks
• Adversarial Attacks on Linear Contextual Bandits
• Discrete Adversarial Attacks and Submodular Optimization with Applications to Text Classification
• SAT: Improving Adversarial Training via Curriculum-Based Loss Smoothing
• Adversarial Defense by Latent Style Transformations
• Improving adversarial robustness of deep neural networks by using semantic information
• Denoising Distant Supervision for Relation Extraction via Instance-Level Adversarial Training
• Enhancing the Robustness of Deep Neural Networks by Boundary Conditional GAN
• ComDefend: An Efficient Image Compression Model to Defend Adversarial Examples
• InfoAT: Improving Adversarial Training Using the Information Bottleneck Principle
• Node Injection for Class-specific Network Poisoning
• Towards Assessing the Synthetic-to-Measured Adversarial Vulnerability of SAR ATR
• Unifying Adversarial Training Algorithms with Flexible Deep Data Gradient Regularization
• SPAA: Stealthy Projector-based Adversarial Attacks on Deep Image Classifiers
• Detect & Reject for Transferability of Black-box Adversarial Attacks Against Network Intrusion Detection Systems
• Towards Analyzing Semantic Robustness of Deep Neural Networks
• A scalable saliency-based Feature selection method with instance level information
• CheXphoto: 10,000+ Photos and Transformations of Chest X-rays for Benchmarking Deep Learning Robustness
• Potential adversarial samples for white-box attacks
• Imperceptible Adversarial Examples by Spatial Chroma-Shift
• Universum-inspired Supervised Contrastive Learning
• Metamorphic Detection of Adversarial Examples in Deep Learning Models With Affine Transformations
• Should Adversarial Attacks Use Pixel p-Norm?
• Simple and Efficient Partial Graph Adversarial Attack: A New Perspective
• FedCSD: A Federated Learning Based Approach for Code-Smell Detection
• Feature Perturbation Augmentation for Reliable Evaluation of Importance Estimators in Neural Networks
• Does CLIP Know My Face?
• Simulator-based explanation and debugging of hazard-triggering events in DNN-based safety-critical systems
• Deeply Explain CNN via Hierarchical Decomposition
• Adversarial Attack for Uncertainty Estimation: Identifying Critical Regions in Neural Networks
• A review of machine learning in processing remote sensing data for mineral exploration
• Single-Node Attacks for Fooling Graph Neural Networks
• SenSeI: Sensitive Set Invariance for Enforcing Individual Fairness
• Enhancing Robustness of Deep Neural Networks Against Adversarial Malware Samples: Principles, Framework, and AICS'2019 Challenge
• Achieving Adversarial Robustness via Sparsity
• Temporal-Clustering Invariance in Irregular Healthcare Time Series
• Solving smooth min-min and min-max problems by mixed oracle algorithms
• Deep Learning for Predictive Analytics in Reversible Steganography
• Adversarial Margin Maximization Networks
• SADA: Semantic Adversarial Diagnostic Attacks for Autonomous Applications
• EigenRank by Committee: A Data Subset Selection and Failure Prediction paradigm for Robust Deep Learning based Medical Image Segmentation
• Action Shuffling for Weakly Supervised Temporal Localization
• Stochastic Zeroth-order Optimization via Variance Reduction method
• Towards Understanding the Faults of JavaScript-Based Deep Learning Systems
• Physical Adversarial Attacks on Deep Neural Networks for Traffic Sign Recognition: A Feasibility Study
• Detecting Backdoors in Neural Networks Using Novel Feature-Based Anomaly Detection
• Falsification-Based Robust Adversarial Reinforcement Learning
• Position-Aware Tagging for Aspect Sentiment Triplet Extraction
• Deep learning for nano-photonic materials -- The solution to everything!?
• A Little Fog for a Large Turn
• Robust Image Classification Using A Low-Pass Activation Function and DCT Augmentation
• Attacking Graph-based Classification via Manipulating the Graph Structure
• Adversarial Sampling for Fairness Testing in Deep Neural Network
• Robustifying Models Against Adversarial Attacks by Langevin Dynamics
• Two Souls in an Adversarial Image: Towards Universal Adversarial Example Detection using Multi-view Inconsistency
• Zeroth-Order Algorithms for Smooth Saddle-Point Problems
• A Survey of AI-Related Cyber Security Risks and Countermeasures in Mobility-as-a-Service
• Adversarial speech for voice privacy protection from Personalized Speech generation
• Dual Attention Suppression Attack: Generate Adversarial Camouflage in Physical World
• UGAN: Untraceable GAN for Multi-Domain Face Translation
• ContextMix: A context-aware data augmentation method for industrial visual inspection systems
• Robustness Testing for Multi-Agent Reinforcement Learning: State Perturbations on Critical Agents
• Improving Robustness and Uncertainty Modelling in Neural Ordinary Differential Equations
• Treatment Learning Causal Transformer for Noisy Image Classification
• Adversarial Image Color Transformations in Explicit Color Filter Space
• DS-UI: Dual-Supervised Mixture of Gaussian Mixture Models for Uncertainty Inference
• Adversarial attacks hidden in plain sight
• Certified Robustness to Word Substitution Ranking Attack for Neural Ranking Models
• Few-shot Forgery Detection via Guided Adversarial Interpolation
• Interpretability is a Kind of Safety: An Interpreter-based Ensemble for Adversary Defense
• A Thorough Comparison Study on Adversarial Attacks and Defenses for Common Thorax Disease Classification in Chest X-rays
• Generating Image Adversarial Examples by Embedding Digital Watermarks
• UnbiasedNets: A Dataset Diversification Framework for Robustness Bias Alleviation in Neural Networks
• Bridging the Performance Gap between FGSM and PGD Adversarial Training
• Perturbation Analysis of Gradient-based Adversarial Attacks
• Defence against adversarial attacks using classical and quantum-enhanced Boltzmann machines
• Entity Recognition from Colloquial Text
• How benign is benign overfitting?
• Efficient ConvNets for Analog Arrays
• Attention Allocation Aid for Visual Search
• Backdoor Attacks on Crowd Counting
• Bayesian Neural Networks for Reversible Steganography
• On the Limitation of MagNet Defense against $L_1$-based Adversarial Examples
• A Robust Adversarial Network-Based End-to-End Communications System With Strong Generalization Ability Against Adversarial Attacks
• Subverting Fair Image Search with Generative Adversarial Perturbations
• Appending Adversarial Frames for Universal Video Attack
• ATHENA: A Framework based on Diverse Weak Defenses for Building Adversarial Defense
• Manifold Regularization for Locally Stable Deep Neural Networks
• In Search of netUnicorn: A Data-Collection Platform to Develop Generalizable ML Models for Network Security Problems
• Exploring the Robustness of NMT Systems to Nonsensical Inputs
• Non-parametric estimation of Jensen-Shannon Divergence in Generative Adversarial Network training
• HDXplore: Automated Blackbox Testing of Brain-Inspired Hyperdimensional Computing
• Robust Node Classification on Graphs: Jointly from Bayesian Label Transition and Topology-based Label Propagation
• Spatio-Temporal Perturbations for Video Attribution
• Perturbing Attention Gives You More Bang for the Buck: Subtle Imaging Perturbations That Efficiently Fool Customized Diffusion Models
• ImpNet: Imperceptible and blackbox-undetectable backdoors in compiled neural networks
• Eight challenges in developing theory of intelligence
• EREBA: Black-box Energy Testing of Adaptive Neural Networks
• Defense against adversarial attacks on deep convolutional neural networks through nonlocal denoising
• Geometric robustness of deep networks: analysis and improvement
• Effects of Forward Error Correction on Communications Aware Evasion Attacks
• Hijack Vertical Federated Learning Models As One Party
• Universal Lipschitz Approximation in Bounded Depth Neural Networks
• Towards Efficiently Evaluating the Robustness of Deep Neural Networks in IoT Systems: A GAN-based Method
• Sparse Adversarial Video Attacks with Spatial Transformations
• Optimal Attack and Defense for Reinforcement Learning
• Gradient-Based Adversarial Training on Transformer Networks for Detecting Check-Worthy Factual Claims
• Towards Robust GAN-generated Image Detection: a Multi-view Completion Representation
• Block Switching: A Stochastic Approach for Deep Learning Security
• Adversarial Robustness by Design through Analog Computing and Synthetic Gradients
• Towards Effective and Efficient Adversarial Defense with Diffusion Models for Robust Visual Tracking
• Robustness of Deep Neural Networks for Micro-Doppler Radar Classification
• Certifiably-Robust Federated Adversarial Learning via Randomized Smoothing
• Optimal Transport as a Defense Against Adversarial Attacks
• A Visual Analytics Framework for Adversarial Text Generation
• Adversarial Machine Learning Phases of Matter
• On the Detection of Adaptive Adversarial Attacks in Speaker Verification Systems
• Using Undervolting as an On-Device Defense Against Adversarial Machine Learning Attacks
• Understanding the Robustness of Skeleton-based Action Recognition under Adversarial Attack
• LAMPAT: Low-Rank Adaption for Multilingual Paraphrasing Using Adversarial Training
• CounteRGAN: Generating Realistic Counterfactuals with Residual Generative Adversarial Nets
• On the Resilience of Machine Learning-Based IDS for Automotive Networks
• A multi-component framework for the analysis and design of explainable artificial intelligence
• Targeted Nonlinear Adversarial Perturbations in Images and Videos
• Ulixes: Facial Recognition Privacy with Adversarial Machine Learning
• Effective and Robust Detection of Adversarial Examples via Benford-Fourier Coefficients
• DeepLens: Interactive Out-of-distribution Data Detection in NLP Models
• Boosting Adversarial Attacks on Neural Networks with Better Optimizer
• Adversarial Robustness Guarantees for Classification with Gaussian Processes
• Adversarial Attacks on Monocular Pose Estimation
• A Survey on Out-of-Distribution Evaluation of Neural NLP Models
• Towards Logical Specification of Statistical Machine Learning
• Denoised Internal Models: a Brain-Inspired Autoencoder against Adversarial Attacks
• Improving the Transferability of Adversarial Attacks on Face Recognition with Beneficial Perturbation Feature Augmentation
• Adversarial jamming attacks and defense strategies via adaptive deep reinforcement learning
• A Generic Self-Supervised Framework of Learning Invariant Discriminative Features
• Generating Universal Adversarial Perturbations for Quantum Classifiers
• Entropy Guided Adversarial Model for Weakly Supervised Object Localization
• APRICOT: A Dataset of Physical Adversarial Attacks on Object Detection
• Laplacian Networks: Bounding Indicator Function Smoothness for Neural Network Robustness
• Trust Region Based Adversarial Attack on Neural Networks
• Mitigating Evasion Attacks to Deep Neural Networks via Region-based Classification
• How Robust Are Character-Based Word Embeddings in Tagging and MT Against Wrod Scramlbing or Randdm Nouse?
• Analysis of Explainable Artificial Intelligence Methods on Medical Image Classification
• Synthesizing Action Sequences for Modifying Model Decisions
• Robustness of Visual Explanations to Common Data Augmentation
• LiBRe: A Practical Bayesian Approach to Adversarial Detection
• SoK: Systematic analysis of adversarial threats against deep learning approaches for autonomous anomaly detection systems in SDN-IoT networks
• Adversarial Defense of Image Classification Using a Variational Auto-Encoder
• Frequency Centric Defense Mechanisms against Adversarial Examples
• Adversarial Security Attacks and Perturbations on Machine Learning and Deep Learning Methods
• CNNs Avoid Curse of Dimensionality by Learning on Patches
• Imbalanced Gradients: A Subtle Cause of Overestimated Adversarial Robustness
• Headless Horseman: Adversarial Attacks on Transfer Learning Models
• Automatic Tracking of the Muscle Tendon Junction in Healthy and Impaired Subjects using Deep Learning
• On Data Augmentation and Adversarial Risk: An Empirical Analysis
• Manipulating hidden-Markov-model inferences by corrupting batch data
• Adversarial Artifact Detection in EEG-Based Brain-Computer Interfaces
• Adversarial Robustness through the Lens of Convolutional Filters
• Probing the Purview of Neural Networks via Gradient Analysis
• Relationship between manifold smoothness and adversarial vulnerability in deep learning with local errors
• GAP++: Learning to generate target-conditioned adversarial examples
• Adversarial Attacks in a Multi-view Setting: An Empirical Study of the Adversarial Patches Inter-view Transferability
• Adversarial Exposure Attack on Diabetic Retinopathy Imagery Grading
• Consistent Counterfactuals for Deep Models
• The Intriguing Relation Between Counterfactual Explanations and Adversarial Examples
• Improving the Robustness of Transformer-based Large Language Models with Dynamic Attention
• On Attribution of Deepfakes
• SOAR: Second-Order Adversarial Regularization
• Security Consideration For Deep Learning-Based Image Forensics
• Adversarial Attack and Defense in Deep Ranking
• Sharp Statistical Guarantees for Adversarially Robust Gaussian Classification
• RoVISQ: Reduction of Video Service Quality via Adversarial Attacks on Deep Learning-based Video Compression
• Benchmarking Spiking Neural Network Learning Methods with Varying Locality
• Robust Proxy: Improving Adversarial Robustness by Robust Proxy Learning
• Improved Detection of Adversarial Images Using Deep Neural Networks
• Attacking and Defending Machine Learning Applications of Public Cloud
• Black-box Gradient Attack on Graph Neural Networks: Deeper Insights in Graph-based Attack and Defense
• Adversarial Image Generation by Spatial Transformation in Perceptual Colorspaces
• The Thousand Faces of Explainable AI Along the Machine Learning Life Cycle: Industrial Reality and Current State of Research
• One word at a time: adversarial attacks on retrieval models
• RNAS-CL: Robust Neural Architecture Search by Cross-Layer Knowledge Distillation
• Accessing Passersby Proxemic Signals through a Head-Worn Camera: Opportunities and Limitations for the Blind
• Neural Network Surrogate and Projected Gradient Descent for Fast and Reliable Finite Element Model Calibration: a Case Study on an Intervertebral Disc
• Temporal Shuffling for Defending Deep Action Recognition Models against Adversarial Attacks
• Catch-Up Mix: Catch-Up Class for Struggling Filters in CNN
• Lightweight and Unobtrusive Data Obfuscation at IoT Edge for Remote Inference
• Unbiased deep solvers for linear parametric PDEs
• A Methodology to Identify Cognition Gaps in Visual Recognition Applications Based on Convolutional Neural Networks
• LMD: A Learnable Mask Network to Detect Adversarial Examples for Speaker Verification
• 3D Adversarial Augmentations for Robust Out-of-Domain Predictions
• Adversarial Reinforcement Learning under Partial Observability in Autonomous Computer Network Defence
• On the Robustness of Human Pose Estimation
• Stress Test Evaluation of Biomedical Word Embeddings
• GRIP: Generative Robust Inference and Perception for Semantic Robot Manipulation in Adversarial Environments
• Software Engineering Practices for Machine Learning
• Learning Relation Prototype from Unlabeled Texts for Long-tail Relation Extraction
• SLA${}^2$P: Self-supervised Anomaly Detection with Adversarial Perturbation
• From Shallow to Deep Interactions Between Knowledge Representation, Reasoning and Machine Learning (Kay R. Amel group)
• Black-box Adversarial Attacks on Network-wide Multi-step Traffic State Prediction Models
• Adversarial Parameter Defense by Multi-Step Risk Minimization
• Adversarial Detection: Attacking Object Detection in Real Time
• Investigating Imperceptibility of Adversarial Attacks on Tabular Data: An Empirical Analysis
• Combining Generators of Adversarial Malware Examples to Increase Evasion Rate
• "I know it when I see it". Visualization and Intuitive Interpretability
• Adversarial joint attacks on legged robots
• Black-box Adversarial Attacks on Monocular Depth Estimation Using Evolutionary Multi-objective Optimization
• Towards Defending Multiple ${\ell }_p$-norm Bounded Adversarial Perturbations via Gated Batch Normalization
• Generating Out of Distribution Adversarial Attack using Latent Space Poisoning
• Revisiting Deep Neural Network Test Coverage from the Test Effectiveness Perspective
• Universal Defensive Underpainting Patch: Making Your Text Invisible to Optical Character Recognition
• DiffDefense: Defending against Adversarial Attacks via Diffusion Models
• Revisiting Neural Program Smoothing for Fuzzing
• Universal adversarial perturbations for multiple classification tasks with quantum classifiers
• Assessing Intelligence in Artificial Neural Networks
• Subset Scanning Over Neural Network Activations
• Stochastic Substitute Training: A Gray-box Approach to Craft Adversarial Examples Against Gradient Obfuscation Defenses
• Layerwise Knowledge Extraction from Deep Convolutional Networks
• Adversarial Transferability in Wearable Sensor Systems
• Optimal Feature Manipulation Attacks Against Linear Regression
• Towards an Awareness of Time Series Anomaly Detection Models' Adversarial Vulnerability
• An Orchestrated Empirical Study on Deep Learning Frameworks and Platforms
• A Bayes-Optimal View on Adversarial Examples
• Noisy Computations during Inference: Harmful or Helpful?
• Adversarial Defense by Stratified Convolutional Sparse Coding
• Defense-VAE: A Fast and Accurate Defense against Adversarial Attacks
• Parameter Re-Initialization through Cyclical Batch Size Schedules
• Optimal Transport Classifier: Defending Against Adversarial Attacks by Regularized Deep Embedding
• Scalable Attack on Graph Data by Injecting Vicious Nodes
• Careful What You Wish For: on the Extraction of Adversarially Trained Models
• A Data Augmentation-based Defense Method Against Adversarial Attacks in Neural Networks
• Updating Windows Malware Detectors: Balancing Robustness and Regression against Adversarial EXEmples
• MagDR: Mask-guided Detection and Reconstruction for Defending Deepfakes
• Privacy Inference Attacks and Defenses in Cloud-based Deep Neural Network: A Survey
• Leveraging Systematic Knowledge of 2D Transformations
• AutoGAN: Robust Classifier Against Adversarial Attacks
• kk2018 at SemEval-2020 Task 9: Adversarial Training for Code-Mixing Sentiment Classification
• Synthesis-guided Adversarial Scenario Generation for Gray-box Feedback Control Systems with Sensing Imperfections
• AccelAT: A Framework for Accelerating the Adversarial Training of Deep Neural Networks through Accuracy Gradient
• Gradient descent in Gaussian random fields as a toy model for high-dimensional optimisation in deep learning
• Sometimes We Want Translationese
• 3D Point Cloud Completion with Geometric-Aware Adversarial Augmentation
• Intermediate Level Adversarial Attack for Enhanced Transferability
• Sequential Randomized Smoothing for Adversarially Robust Speech Recognition
• Trust but Verify: An Information-Theoretic Explanation for the Adversarial Fragility of Machine Learning Systems, and a General Defense against Adversarial Attacks
• RamBoAttack: A Robust Query Efficient Deep Neural Network Decision Exploit
• Advancing Adversarial Training by Injecting Booster Signal
• Extending Neural Network Verification to a Larger Family of Piece-wise Linear Activation Functions
• mFI-PSO: A Flexible and Effective Method in Adversarial Image Generation for Deep Neural Networks
• Is Face Recognition Safe from Realizable Attacks?
• On the relationship between class selectivity, dimensionality, and robustness
• Automatic Feature Fairness in Recommendation via Adversaries
• Pelta: Shielding Transformers to Mitigate Evasion Attacks in Federated Learning
• Stochastic Combinatorial Ensembles for Defending Against Adversarial Examples
• Automation of reversible steganographic coding with nonlinear discrete optimisation
• Virtual Adversarial Ladder Networks For Semi-supervised Learning
• Benford's law: what does it say on adversarial images?
• Neural Generators of Sparse Local Linear Models for Achieving both Accuracy and Interpretability
• The Power of Typed Affine Decision Structures: A Case Study
• Capsule network with shortcut routing
• Resilient Linear Classification: An Approach to Deal with Attacks on Training Data
• Damage GAN: A Generative Model for Imbalanced Data
• Generative adversarial network based on chaotic time series
• Adversarial Attacks on Classifiers for Eye-based User Modelling
• A Reflection on Learning from Data: Epistemology Issues and Limitations
• Mitigating Gradient-based Adversarial Attacks via Denoising and Compression
• Amplification trojan network: Attack deep neural networks by amplifying their inherent weakness
• TACO: Adversarial Camouflage Optimization on Trucks to Fool Object Detectors
• Weight Map Layer for Noise and Adversarial Attack Robustness
• Adversarial Transformations for Semi-Supervised Learning
• Why Accuracy Is Not Enough: The Need for Consistency in Object Detection
• Deep reinforcement learning-based image classification achieves perfect testing set accuracy for MRI brain tumors with a training set of only 30 images
• Minimax Defense against Gradient-based Adversarial Attacks
• A Novel User Representation Paradigm for Making Personalized Candidate Retrieval
• Variation Enhanced Attacks Against RRAM-based Neuromorphic Computing System
• A Singular Value Perspective on Model Robustness
• Designing Adversarially Resilient Classifiers using Resilient Feature Engineering
• A Self-supervised Approach for Adversarial Robustness
• A Survey on Trust Metrics for Autonomous Robotic Systems
• Adversarial Robustness in Deep Learning: Attacks on Fragile Neurons
• Improving Model Robustness with Latent Distribution Locally and Globally
• Analytical Moment Regularizer for Gaussian Robust Networks
• Good-looking but Lacking Faithfulness: Understanding Local Explanation Methods through Trend-based Testing
• Latent Adversarial Defence with Boundary-guided Generation
• Soft Adversarial Training Can Retain Natural Accuracy
• torchosr -- a PyTorch extension package for Open Set Recognition models evaluation in Python
• On the predictability in reversible steganography
• Adequacy of the Gradient-Descent Method for Classifier Evasion Attacks
• Beyond Neural-on-Neural Approaches to Speaker Gender Protection
• Language Guided Adversarial Purification
• Generalization Error Analysis of Neural networks with Gradient Based Regularization
• AICAttack: Adversarial Image Captioning Attack with Attention-Based Optimization
• Can Perceptual Guidance Lead to Semantically Explainable Adversarial Perturbations?
• A Robust Classification-autoencoder to Defend Outliers and Adversaries
• Iterative Window Mean Filter: Thwarting Diffusion-based Adversarial Purification
• A Unified Game-Theoretic Interpretation of Adversarial Robustness
• Adversarial attacks to image classification systems using evolutionary algorithms
• MEAT: Median-Ensemble Adversarial Training for Improving Robustness and Generalization
• Generating Adversarial Disturbances for Controller Verification
• Coverage-Guaranteed Prediction Sets for Out-of-Distribution Data
• Toward Few-step Adversarial Training from a Frequency Perspective
• Lost In Translation: Generating Adversarial Examples Robust to Round-Trip Translation
• Towards User Guided Actionable Recourse
• Combining Two Adversarial Attacks Against Person Re-Identification Systems
• A Geometrical Approach to Evaluate the Adversarial Robustness of Deep Neural Networks
• Amoeba: Circumventing ML-supported Network Censorship via Adversarial Reinforcement Learning
• XploreNAS: Explore Adversarially Robust & Hardware-efficient Neural Architectures for Non-ideal Xbars
• To Make Yourself Invisible with Adversarial Semantic Contours
• Adversarial Monte Carlo Meta-Learning of Optimal Prediction Procedures
• Towards Robust Off-Policy Evaluation via Human Inputs
• Negative-ResNet: Noisy Ambulatory Electrocardiogram Signal Classification Scheme
• Residual Error: a New Performance Measure for Adversarial Robustness
• Adversarial Black-Box Attacks On Text Classifiers Using Multi-Objective Genetic Optimization Guided By Deep Networks
• Robustness of on-device Models: Adversarial Attack to Deep Learning Models on Android Apps
• Statistical inference for individual fairness
• Gradient-based Adversarial Deep Modulation Classification with Data-driven Subsampling
• Exploring Model Robustness with Adaptive Networks and Improved Adversarial Training
• Learning Discriminators as Energy Networks in Adversarial Learning
• Reinforcement Learning for Autonomous Defence in Software-Defined Networking
• An improved cosmological parameter inference scheme motivated by deep learning
• Attacking Optical Flow
• Spot Evasion Attacks: Adversarial Examples for License Plate Recognition Systems with Convolutional Neural Networks
• Representation of White- and Black-Box Adversarial Examples in Deep Neural Networks and Humans: A Functional Magnetic Resonance Imaging Study
• Kernelized Capsule Networks
• Multi-head Ensemble of Smoothed Classifiers for Certified Robustness
• Improved Surrogates in Inertial Confinement Fusion with Manifold and Cycle Consistencies
• FoveaTer: Foveated Transformer for Image Classification
• Analytical bounds on the local Lipschitz constants of affine-ReLU functions
• Scope and Sense of Explainability for AI-Systems
• Chameleon: A Semi-AutoML framework targeting quick and scalable development and deployment of production-ready ML systems for SMEs
• Unique Exams: Designing assessments for integrity and fairness
• Minimal Sufficient Views: A DNN model making predictions with more evidence has higher accuracy
• Blind Adversarial Pruning: Balance Accuracy, Efficiency and Robustness
• Why Blocking Targeted Adversarial Perturbations Impairs the Ability to Learn
• Who is Responsible for Adversarial Defense?
• An Adversarial Approach for Explainable AI in Intrusion Detection Systems
• Mapper Based Classifier
• When Side-Channel Attacks Break the Black-Box Property of Embedded Artificial Intelligence
• Adversarial AutoMixup
• Robust and Information-theoretically Safe Bias Classifier against Adversarial Attacks
• Enhance Diffusion to Improve Robust Generalization
• Adversarial Robustness on Image Classification with $k$-means
• Robust Network Slicing: Multi-Agent Policies, Adversarial Attacks, and Defensive Strategies
• Neural Belief Reasoner
• Adversarial Edit Attacks for Tree Data
• Addressing Weak Decision Boundaries in Image Classification by Leveraging Web Search and Generative Models
• Natural Adversarial Objects
• Adversarial training applied to Convolutional Neural Network for photometric redshift predictions
• Search Space of Adversarial Perturbations against Image Filters
• Worrisome Properties of Neural Network Controllers and Their Symbolic Representations
• Predicting "Design Gaps" in the Market: Deep Consumer Choice Models under Probabilistic Design Constraints
• Global Robustness Verification Networks
• Augmenting Model Robustness with Transformation-Invariant Attacks
• An Information-Theoretic Explanation for the Adversarial Fragility of AI Classifiers
• Out-of-Sample Testing for GANs
• Sentiment Tagging with Partial Labels using Modular Architectures
• Wasserstein Introspective Neural Networks
• An Evasion Attack against Stacked Capsule Autoencoder
• Towards Precise Observations of Neural Model Robustness in Classification
• Can I teach a robot to replicate a line art
• Robust Single-step Adversarial Training with Regularizer
• Dynamic Black-box Backdoor Attacks on IoT Sensory Data
• Comparing Reinforcement Learning and Human Learning using the Game of Hidden Rules
• Attack Assessment and Augmented Identity Recognition for Human Skeleton Data
• Access Control with Encrypted Feature Maps for Object Detection Models
• Generating Adversarial Inputs Using A Black-box Differential Technique
• Detecting Adversarial Samples Using Density Ratio Estimates
• Unveiling Hidden Vulnerabilities in Digital Human Generation via Adversarial Attacks
• Adversarial Jamming for a More Effective Constellation Attack
• Robust Partial-Label Learning by Leveraging Class Activation Values
• Exposing Semantic Segmentation Failures via Maximum Discrepancy Competition
• Bounding The Number of Linear Regions in Local Area for Neural Networks with ReLU Activations
• MixDefense: A Defense-in-Depth Framework for Adversarial Example Detection Based on Statistical and Semantic Analysis
• Non-Determinism in Neural Networks for Adversarial Robustness
• Scalable Multi-Class Bayesian Support Vector Machines for Structured and Unstructured Data
• Deep image prior inpainting of ancient frescoes in the Mediterranean Alpine arc
• Can audio-visual integration strengthen robustness under multimodal attacks?
• Cross-domain Cross-architecture Black-box Attacks on Fine-tuned Models with Transferred Evolutionary Strategies
• A Study on the Uncertainty of Convolutional Layers in Deep Neural Networks
• The Compact Support Neural Network
• Versatile Verification of Tree Ensembles
• ATRAS: Adversarially Trained Robust Architecture Search
• Heat and Blur: An Effective and Fast Defense Against Adversarial Examples
• Residual Networks as Nonlinear Systems: Stability Analysis using Linearization
• Adversarial Defenses via Vector Quantization
• Generating Minimal Adversarial Perturbations with Integrated Adaptive Gradients
• Thwarting finite difference adversarial attacks with output randomization
• Amicable Aid: Perturbing Images to Improve Classification Performance
• Towards Optimal Randomized Strategies in Adversarial Example Game
• Adversarial attacks on neural networks through canonical Riemannian foliations
• State-Reification Networks: Improving Generalization by Modeling the Distribution of Hidden Representations
• Efficient Low-Rank Semidefinite Programming with Robust Loss Functions
• Perception Matters: Exploring Imperceptible and Transferable Anti-forensics for GAN-generated Fake Face Imagery Detection
• Towards A Conceptually Simple Defensive Approach for Few-shot classifiers Against Adversarial Support Samples
• Spatially Correlated Patterns in Adversarial Images
• An Efficient Pre-processing Method to Eliminate Adversarial Effects
• Trust but Verify: Assigning Prediction Credibility by Counterfactual Constrained Learning
• Calibrated Adversarial Training
• Mixing between the Cross Entropy and the Expectation Loss Terms
• Machine learning pipeline for battery state of health estimation
• PredCoin: Defense against Query-based Hard-label Attack
• Extremal learning: extremizing the output of a neural network in regression problems
• Interpreting Attributions and Interactions of Adversarial Attacks
• advPattern: Physical-World Attacks on Deep Person Re-Identification via Adversarially Transformable Patterns
• Audio Adversarial Examples: Attacks Using Vocal Masks
• What classifiers know what they don't?
• Provable Generalization of SGD-trained Neural Networks of Any Width in the Presence of Adversarial Label Noise
• Understanding Robustness in Teacher-Student Setting: A New Perspective
• Minimum sharpness: Scale-invariant parameter-robustness of neural networks
• Group-Structured Adversarial Training
• Adversarial Attacks on Deep Models for Financial Transaction Records
• On educating machines
• An Adaptive Empirical Bayesian Method for Sparse Deep Learning
• A Primer on Multi-Neuron Relaxation-based Adversarial Robustness Certification
• ODE guided Neural Data Augmentation Techniques for Time Series Data and its Benefits on Robustness
• A Comparison of Methods for Neural Network Aggregation
• On Model Robustness Against Adversarial Examples
• Boosting Black-Box Adversarial Attacks with Meta Learning
• Beyond Categorical Label Representations for Image Classification
• Simpler Certified Radius Maximization by Propagating Covariances
• Training Efficiency and Robustness in Deep Learning
• Robustness to Programmable String Transformations via Augmented Abstract Training
• Regularized adversarial examples for model interpretability
• Robustness Tests of NLP Machine Learning Models: Search and Semantically Replace
• Boosting the Robustness Verification of DNN by Identifying the Achilles's Heel
• Empirical Upper Bound, Error Diagnosis and Invariance Analysis of Modern Object Detectors
• FUNN: Flexible Unsupervised Neural Network
• Do Deep Minds Think Alike? Selective Adversarial Attacks for Fine-Grained Manipulation of Multiple Deep Neural Networks
• Strategies for Robust Image Classification
• Spatiotemporal Attacks for Embodied Agents
• Adversarial Perturbation Intensity Achieving Chosen Intra-Technique Transferability Level for Logistic Regression
• Efficient Adversarial Input Generation via Neural Net Patching
• Formal Verification of Long Short-Term Memory based Audio Classifiers: A Star based Approach
• Linking average- and worst-case perturbation robustness via class selectivity and dimensionality
• Affine-Invariant Robust Training
• Efficient and Transferable Adversarial Examples from Bayesian Neural Networks
• Nested Learning For Multi-Granular Tasks
• Derivation of Information-Theoretically Optimal Adversarial Attacks with Applications to Robust Machine Learning
• Finding Closure: A Closer Look at the Gestalt Law of Closure in Convolutional Neural Networks
• Structured Ensembles: an Approach to Reduce the Memory Footprint of Ensemble Methods
• FireBERT: Hardening BERT-based classifiers against adversarial attack
• Semantic Network Interpretation
• BAARD: Blocking Adversarial Examples by Testing for Applicability, Reliability and Decidability
• Towards Explaining Adversarial Examples Phenomenon in Artificial Neural Networks
• When Automatic Voice Disguise Meets Automatic Speaker Verification
• Group-based Fair Learning Leads to Counter-intuitive Predictions
• Adversarial Detection by Approximation of Ensemble Boundary
• Closing the Gap: Achieving Better Accuracy-Robustness Tradeoffs against Query-Based Attacks
• Global Optimum Search in Quantum Deep Learning
• Adversarial VC-dimension and Sample Complexity of Neural Networks
• P-CapsNets: a General Form of Convolutional Neural Networks
• Guess First to Enable Better Compression and Adversarial Robustness
• Noisy Feature Mixup
• Generating Adversarial Perturbation with Root Mean Square Gradient
• Can Intelligent Hyperparameter Selection Improve Resistance to Adversarial Examples?
• Local Linearity and Double Descent in Catastrophic Overfitting
• Anomaly-Aware Semantic Segmentation by Leveraging Synthetic-Unknown Data
• Familiarity-Based Open-Set Recognition Under Adversarial Attacks
• Improved Algorithms for White-Box Adversarial Streams