Article

Beyond AI: Fooling DNN Model in Physical World

2024-11-13

Abstract excerpt

<title>Abstract</title> <p>Deep neural networks (DNNs) are becoming increasingly crucial across various fields, but their wide adoption raises critical security concerns. This paper proposes an attack that discreetly manipulates model predictions by strategically embedding a tiny patch into the input image. This adversarial patch remains imperceptible to human observers yet effectively compels the model to miscla...

Topics

Open a Topic to create a Post that cites this publication.

Identifiers and source

Literature Corpus work: 0bb40c37-6987-5967-8f23-ba352affd64e
DOI: 10.21203/rs.3.rs-5352905/v1

Open publication

Related research

Semantic proximity does not establish scientific evidence.

Click a neighbor to travelStep 1 · 12 closest

Beyond AI: Fooling DNN Model in Physical WorldDOI 10.21203/rs.3.rs-5352905/v1Read

Generation and Countermeasures of Adversarial Examples on Vision: A SurveySemantic proximity Read

How Adversarial attacks affect Deep Neural Networks Detecting COVID-19?Semantic proximity Read

Transient Deception: Recent Advances and Trends in Short-Lived Adversarial ExamplesSemantic proximity Read

Image Synthesis-based Backdoor Attack Approach in the Face Classification TaskSemantic proximity Read

Improving Transferability of Adversarial Examples with Mixed-Representation AttackSemantic proximity Read

Imperceptible Adversarial Attacks against Traffic Scene RecognitionSemantic proximity Read

Color-channel Adversarial Attack with Resolution based CamouflagingSemantic proximity Read

Natural Images Allow Universal Adversarial Attacks on Medical Image Classification Using Deep Neural Networks With Transfer LearningSemantic proximity Read

Chained Dual-GAN: A Generalized Defense against Adversarial AttacksSemantic proximity Read

Functionality-preserving Adversarial Machine Learning For Robust Classification In Cybersecurity And Intrusion Detection Domains: A SurveySemantic proximity Read

Detection of Attacks with An Adversarial Machine Learning ApproachSemantic proximity Read

Simulating a Primary Visual Cortex at the Front of CNNs Improves Robustness to Image PerturbationsSemantic proximity Read