mlsanityData to 14 Oct 2026

Hide and Seek: on the Stealthiness of Attacks against Deep Learning Systems

arXiv 2205.15944 ↗First posted 31 May 2022cs.CR cs.AI cs.LGcited by 3

Not on the map. Why?

Abstract (arXiv)

With the growing popularity of artificial intelligence and machine learning, a wide spectrum of attacks against deep learning models have been proposed in the literature. Both the evasion attacks and the poisoning attacks attempt to utilize adversarially altered samples to fool the victim model to misclassify the adversarial sample. While such attacks claim to be or are expected to be stealthy, i.e., imperceptible to human eyes, such claims are rarely evaluated. In this paper, we present the first large-scale study on the stealthiness of adversarial samples used in the attacks against deep learning. We have implemented 20 representative adversarial ML attacks on six popular benchmarking datasets. We evaluate the stealthiness of the attack samples using two complementary approaches: (1) a numerical study that adopts 24 metrics for image similarity or quality assessment; and (2) a user study of 3 sets of questionnaires that has collected 20,000+ annotations from 1,000+ responses. Our results show that the majority of the existing attacks introduce nonnegligible perturbations that are not stealthy to human eyes. We further analyze the factors that contribute to attack stealthiness. We further examine the correlation between the numerical analysis and the user studies, and demonstrate that some image quality metrics may provide useful guidance in attack designs, while there is still a significant gap between assessed image quality and visual stealthiness of attacks.

Key facts. Cited by 3 papers in the map (data through 14 Oct 2026). Builds on 28 of its 51 references that are in the map. First posted on arXiv 31 May 2022.

Builds on

28 of 51 references.

References · 28

  1. 2018The Unreasonable Effectiveness of Deep Features as a Perceptual Metriccited by 11,177
  2. 2014ImageNet Large Scale Visual Recognition Challengecited by 8,498
  3. 2014Explaining and Harnessing Adversarial Examplescited by 7,364
  4. 2017Towards Deep Learning Models Resistant to Adversarial Attackscited by 6,415
  5. 2015Domain-Adversarial Training of Neural Networkscited by 3,044
  6. 2016Towards Evaluating the Robustness of Neural Networkscited by 2,998
  7. 2014Unsupervised Domain Adaptation by Backpropagationcited by 1,917
  8. 2016Adversarial examples in the physical worldcited by 1,769
  9. 2020Reliable evaluation of adversarial robustness with an ensemble of diverse parameter-free attackscited by 1,451
  10. 2015DeepFool: a simple and accurate method to fool deep neural networkscited by 1,318
  11. 2017BadNets: Identifying Vulnerabilities in the Machine Learning Model Supply Chaincited by 1,106
  12. 2017Targeted Backdoor Attacks on Deep Learning Systems Using Data Poisoningcited by 1,055
16 more
  1. 2020Fast is better than free: Revisiting adversarial trainingcited by 727
  2. 2021Pervasive Label Errors in Test Sets Destabilize Machine Learning Benchmarkscited by 395
  3. 2019Hidden Trigger Backdoor Attackscited by 371
  4. 2018Adversarial Attacks and Defences: A Surveycited by 259
  5. 2019A new Backdoor Attack in CNNs by training set corruption without label poisoningcited by 242
  6. 2020Focal Frequency Loss for Image Reconstruction and Synthesiscited by 206
  7. 2021Rethinking the Backdoor Attacks' Triggers: A Frequency Perspectivecited by 161
  8. 2018Backdoor Embedding in Convolutional Neural Network Models via Invisible Perturbationcited by 134
  9. 2019TABOR: A Highly Accurate Approach to Inspecting and Restoring Trojan Backdoors in AI Systemscited by 122
  10. 2020An Embarrassingly Simple Approach for Trojan Attack in Deep Neural Networkscited by 97
  11. 2018ExpandNet: A Deep Convolutional Neural Network for High Dynamic Range Expansion from Low Dynamic Range Contentcited by 89
  12. 2018Semantic Adversarial Examplescited by 86
  13. 2018Backdooring Convolutional Neural Networks via Targeted Weight Perturbationscited by 69
  14. 2021DeepPayload: Black-box Backdoor Attack on Deep Learning Models through Neural Payload Injectioncited by 44
  15. 2020AIM 2020 Challenge on Real Image Super-Resolution: Methods and Resultscited by 17
  16. 2019SmoothFool: An Efficient Framework for Computing Smooth Adversarial Perturbationscited by 12

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 2; 2026: 1.

3 papers.

  1. 2025How stealthy is stealthy? Studying the Efficacy of Black-Box Adversarial Attacks in the Real Worldcited by 1
  2. 2025CatBack: Universal Backdoor Attacks on Tabular Data via Categorical Encodingcited by 0
  3. 2026Quiet Triggers, Loud Footprints: A Tri-Space Measurement Study of Backdoor Stealthinesscited by 0