mlsanityData to 7 Oct 2026

NO Need to Worry about Adversarial Examples in Object Detection in Autonomous Vehicles

arXiv 1707.03501 ↗First posted 12 Jul 2017cs.CV cs.AI cs.CRcited by 48

Not on the map. Why?

Abstract (arXiv)

It has been shown that most machine learning algorithms are susceptible to adversarial perturbations. Slightly perturbing an image in a carefully chosen direction in the image space may cause a trained neural network model to misclassify it. Recently, it was shown that physical adversarial examples exist: printing perturbed images then taking pictures of them would still result in misclassification. This raises security and safety concerns. However, these experiments ignore a crucial property of physical objects: the camera can view objects from different distances and at different angles. In this paper, we show experiments that suggest that current constructions of physical adversarial examples do not disrupt object detection from a moving platform. Instead, a trained neural network classifies most of the pictures taken from different distances and angles of a perturbed image correctly. We believe this is because the adversarial property of the perturbation is sensitive to the scale at which the perturbed picture is viewed, so (for example) an autonomous car will misclassify a stop sign only from a small range of distances. Our work raises an important question: can one construct examples that are adversarial for many or most viewing conditions? If so, the construction should offer very significant insights into the internal representation of patterns by deep networks. If not, there is a good prospect that adversarial examples can be reduced to a curiosity with little practical impact.

Key facts. Cited by 48 papers in the map (data through 7 Oct 2026). First posted on arXiv 12 Jul 2017.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 19; 2022: 11; 2023: 8; 2024: 6; 2025: 3; 2026: 1.

48 papers.

Sort:

By area:

  1. 2022Physical Adversarial Attack meets Computer Vision: A Decade Surveycited by 66
  2. 2021Adversarial Example Detection for DNN Models: A Review and Experimental Comparisoncited by 51
  3. 2021Robust Machine Learning Systems: Challenges, Current Trends, Perspectives, and the Road Aheadcited by 40
  4. 2023DAP: A Dynamic Adversarial Patch for Evading Person Detectorscited by 40
  5. 2023Does Physical Adversarial Example Really Matter to Autonomous Driving? Towards System-Level Effect of Adversarial Object Evasion Attackcited by 37
  6. 2021TnT Attacks! Universal Naturalistic Adversarial Patches Against Deep Neural Network Systemscited by 36
  7. 2021Physical Adversarial Attacks on an Aerial Imagery Object Detectorcited by 35
  8. 2022Robust Unlearnable Examples: Protecting Data Against Adversarial Learningcited by 34
  9. 2021Corner Case Generation and Analysis for Safety Assessment of Autonomous Vehiclescited by 27
  10. 2022Fooling the Eyes of Autonomous Vehicles: Robust Physical Adversarial Examples Against Traffic Sign Recognition Systemscited by 27
  11. 2022SoK: On the Semantic AI Security in Autonomous Drivingcited by 22
  12. 2023Physical Adversarial Attacks For Camera-based Smart Systems: Current Trends, Categorization, Applications, Research Challenges, and Future Outlookcited by 20
36 more
  1. 2021"Adversarial Examples" for Proof-of-Learningcited by 19
  2. 2021Learning Coated Adversarial Camouflages for Object Detectorscited by 17
  3. 2022Strong Transferable Adversarial Attacks via Ensembled Asymptotically Normal Distribution Learningcited by 15
  4. 2021PatchCensor: Patch Robustness Certification for Transformers via Exhaustive Testingcited by 11
  5. 2023Dynamic Adversarial Attacks on Autonomous Driving Systemscited by 11
  6. 2023Targeted Adversarial Attacks on Generalizable Neural Radiance Fieldscited by 10
  7. 2021Controlled Caption Generation for Images Through Adversarial Attackscited by 9
  8. 2021Learning Transferable 3D Adversarial Cloaks for Deep Trained Detectorscited by 8
  9. 2022TransCAB: Transferable Clean-Annotation Backdoor to Object Detection with Natural Trigger in Real-Worldcited by 8
  10. 2021IPatch: A Remote Adversarial Patchcited by 7
  11. 2021Attack-agnostic Adversarial Detection on Medical Data Using Explainable Machine Learningcited by 7
  12. 2021Random and Adversarial Bit Error Robustness: Energy-Efficient and Secure DNN Acceleratorscited by 6
  13. 2021Towards Adversarial Patch Analysis and Certified Defense against Crowd Countingcited by 6
  14. 2024Revisiting Physical-World Adversarial Attack on Traffic Sign Recognition: A Commercial Systems Perspectivecited by 6
  15. 2021Simple Post-Training Robustness Using Test Time Augmentations and Random Forestcited by 4
  16. 2022Empirical Evaluation of Physical Adversarial Patch Attacks Against Overhead Object Detection Modelscited by 4
  17. 2024Mitigating Adversarial Perturbations for Deep Reinforcement Learning via Vector Quantizationcited by 4
  18. 2024Natural Language Induced Adversarial Imagescited by 4
  19. 2024Optimizing Multispectral Object Detection: A Bag of Tricks and Comprehensive Benchmarkscited by 4
  20. 2021Messing Up 3D Virtual Environments: Transferable Adversarial 3D Objectscited by 2
  21. 2022The Weaknesses of Adversarial Camouflage in Overhead Imagerycited by 2
  22. 2023Why Don't You Clean Your Glasses? Perception Attacks with Dynamic Optical Perturbationscited by 2
  23. 2021The Effects of Image Distribution and Task on Adversarial Robustnesscited by 1
  24. 2021DriveGuard: Robustification of Automated Driving Systems with Deep Spatio-Temporal Convolutional Autoencodercited by 1
  25. 2022On the Feasibility and Generality of Patch-based Adversarial Attacks on Semantic Segmentation Problemscited by 1
  26. 2022Synthetic Dataset Generation for Adversarial Machine Learning Researchcited by 1
  27. 2022A Human-in-the-Middle Attack against Object Detection Systemscited by 1
  28. 2023ExploreADV: Towards exploratory attack for Neural Networkscited by 1
  29. 2023X-Detect: Explainable Adversarial Patch Detection for Object Detectors in Retailcited by 1
  30. 2025Phi: Preference Hijacking in Multi-modal Large Language Models at Inference Timecited by 1
  31. 2021Improving Robustness with Image Filteringcited by 0
  32. 2024Enhancing Robustness of Indoor Robotic Navigation with Free-Space Segmentation Models Against Adversarial Attackscited by 0
  33. 2024On the Perturbed States for Transformed Input-robust Reinforcement Learningcited by 0
  34. 2025AngleRoCL: Angle-Robust Concept Learning for Physically View-Invariant T2I Adversarial Patchescited by 0
  35. 2025Robust Physical Adversarial Patches Using Dynamically Optimized Clusterscited by 0
  36. 2026Generating Multi-view Adversarial Examples for Visual Geometry Grounded Transformercited by 0