mlsanityData to 13 Oct 2026

AEGIS: A Semantic GAN and Evidential Learning Frameworkfor Robust Adversarial Detection in Vision Sensors

arXiv 2606.28416 ↗First posted 25 Jun 2026cs.CV cs.AIcited by 0

Not on the map. Why?

Abstract (arXiv)

Deep neural networks (DNNs) have shown outstanding performance in visual recognition tasks within vision sensor networks; however, they are still vulnerable to adversarial manipulations and imperceptible perturbations that can lead to erroneous predictions. To address that, this paper presents AEGIS, a semantic aware and uncertainty guided adversarial detection framework designed for robust image classification in vision sensors pipelines. At its core, a SemantiGAN module functions as a multi class semantic discriminator, identifying and filtering visually inconsistent adversarial inputs before they propagate further in the pipeline. For inputs that pass this stage, a stochastic augmentation process generates test time variations, from which handcrafted instability metrics FlipScore, Prediction Inconsistency, Layerwise Cosine Similarity (early and mid layers), and Entropy are computed. These features are aggregated into a compact five dimensional vector and processed by an Evidential Deep Learning (EDL) classifier, which models output evidence using a Dirichlet distribution to yield both class predictions and calibrated uncertainty estimates. Evaluations on the Tiny ImageNet dataset across six categories clean, FGSM, PGD, patch based, functional, and geometric attacks demonstrate the effectiveness of AEGIS. The proposed framework achieves an AUROC of 92.1%, an AUPRC of 90.2%, and an accuracy of 90.7%, outperforming conventional softmax-based detectors in terms of detection performance, robustness, interpretability, and uncertainty calibration.

Key facts. No citing papers in the map yet (data through 13 Oct 2026). Builds on 46 of its 79 references that are in the map. First posted on arXiv 25 Jun 2026.

Builds on

46 of 79 references.

References · 46

  1. 2014Explaining and Harnessing Adversarial Examplescited by 7,336
  2. 2017Towards Deep Learning Models Resistant to Adversarial Attackscited by 6,379
  3. 2013Intriguing properties of neural networkscited by 5,137
  4. 2015Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learningcited by 4,430
  5. 2016Simple and Scalable Predictive Uncertainty Estimation using Deep Ensemblescited by 3,902
  6. 2016Towards Evaluating the Robustness of Neural Networkscited by 2,992
  7. 2016Adversarial examples in the physical worldcited by 1,767
  8. 2018A Simple Unified Framework for Detecting Out-of-Distribution Samples and Adversarial Attackscited by 1,475
  9. 2017Enhancing The Reliability of Out-of-distribution Image Detection in Neural Networkscited by 1,327
  10. 2015The Limitations of Deep Learning in Adversarial Settingscited by 790
  11. 2018Evidential Deep Learning to Quantify Classification Uncertaintycited by 729
  12. 2017Adversarial Patchcited by 516
34 more
  1. 2016Transferability in Machine Learning: from Phenomena to Black-Box Attacks using Adversarial Samplescited by 444
  2. 2018GANomaly: Semi-Supervised Anomaly Detection via Adversarial Trainingcited by 409
  3. 2018Defense-GAN: Protecting Classifiers Against Adversarial Attacks Using Generative Modelscited by 362
  4. 2018Why ReLU networks yield high-confidence predictions far away from the training data and how to mitigate the problemcited by 323
  5. 2018Characterizing Adversarial Subspaces Using Local Intrinsic Dimensionalitycited by 281
  6. 2018Generating Adversarial Examples with Adversarial Networkscited by 258
  7. 2017Detecting Adversarial Samples from Artifactscited by 244
  8. 2018Spatially Transformed Adversarial Examplescited by 195
  9. 2020Robust Pre-Training by Adversarial Contrastive Learningcited by 144
  10. 2018Semantic Adversarial Examplescited by 86
  11. 2016Adversarial Examples Detection in Deep Networks with Convolutional Filter Statisticscited by 77
  12. 2019Semantic Adversarial Attacks: Parametric Transformations That Fool Deep Classifierscited by 56
  13. 2024Good Seed Makes a Good Crop: Discovering Secret Seeds in Text-to-Image Diffusion Modelscited by 48
  14. 2023RLSbench: Domain Adaptation Under Relaxed Label Shiftcited by 30
  15. 2024SceneTAP: Scene-Coherent Typographic Adversarial Planner against Vision-Language Models in Real-World Environmentscited by 22
  16. 2022Benchmarking Adversarial Patch Against Aerial Detectioncited by 21
  17. 2023Mutual-modality Adversarial Attack with Semantic Perturbationcited by 14
  18. 2021Exploiting epistemic uncertainty of the deep learning models to generate adversarial samplescited by 13
  19. 2021Exploiting Multi-Object Relationships for Detecting Adversarial Attacks in Complex Scenescited by 12
  20. 2021When and How to Fool Explainable Models (and Humans) with Adversarial Examplescited by 10
  21. 2024A Structured Review of Literature on Uncertainty in Machine Learning & Deep Learningcited by 10
  22. 2025SegTrans: Transferable Adversarial Examples for Segmentation Modelscited by 8
  23. 2025Bridging Symmetry and Robustness: On the Role of Equivariance in Enhancing Adversarial Robustnesscited by 8
  24. 2021Semantically Stealthy Adversarial Attacks against Segmentation Modelscited by 7
  25. 2022Evidential Deep Learning for Class-Incremental Semantic Segmentationcited by 6
  26. 2024Towards Certification of Uncertainty Calibration under Adversarial Attackscited by 5
  27. 2023Constructing Semantics-Aware Adversarial Examples with a Probabilistic Perspectivecited by 4
  28. 2025LISArD: Learning Image Similarity to Defend Against Gray-box Adversarial Attackscited by 3
  29. 2022Post-selection Inference in Multiverse Analysis (PIMA): an inferential framework based on the sign flipping score testcited by 2
  30. 2022Multiple Perturbation Attack: Attack Pixelwise Under Different ℓp-norms For Better Adversarial Performancecited by 2
  31. 2025Robust Adversarial Quantification via Conflict-Aware Evidential Deep Learningcited by 2
  32. 2025Test-Time Defense Against Adversarial Attacks via Stochastic Resonance of Latent Ensemblescited by 2
  33. 2025AdaGAT: Adaptive Guidance Adversarial Training for the Robustness of Deep Neural Networkscited by 1
  34. 2025DRIFT: Divergent Response in Filtered Transformations for Robust Adversarial Defensecited by 1

Built on by

Not cited yet.