mlsanityData to 7 Oct 2026

Adversarial Active Learning for Deep Networks: a Margin Based Approach

arXiv 1802.09841 ↗First posted 27 Feb 2018cs.LG cs.CV stat.MLcited by 155

Not on the map. Why?

Abstract (arXiv)

We propose a new active learning strategy designed for deep neural networks. The goal is to minimize the number of data annotation queried from an oracle during training. Previous active learning strategies scalable for deep networks were mostly based on uncertain sample selection. In this work, we focus on examples lying close to the decision boundary. Based on theoretical works on margin theory for active learning, we know that such examples may help to considerably decrease the number of annotations. While measuring the exact distance to the decision boundaries is intractable, we propose to rely on adversarial examples. We do not consider anymore them as a threat instead we exploit the information they provide on the distribution of the input space in order to approximate the distance to decision boundaries. We demonstrate empirically that adversarial active queries yield faster convergence of CNNs trained on MNIST, the Shoe-Bag and the Quick-Draw datasets.

Key facts. Cited by 155 papers in the map (data through 7 Oct 2026). First posted on arXiv 27 Feb 2018.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 31; 2022: 25; 2023: 33; 2024: 35; 2025: 13; 2026: 18.

155 papers.

Sort:

By area:

  1. 2021Active Learning by Acquiring Contrastive Examplescited by 154
  2. 2022DeepCore: A Comprehensive Library for Coreset Selection in Deep Learningcited by 146
  3. 2023InfoBatch: Lossless Training Speed Up by Unbiased Dynamic Data Pruningcited by 90
  4. 2021Gone Fishing: Neural Active Learning with Fisher Embeddingscited by 88
  5. 2022A Comparative Survey of Deep Active Learningcited by 88
  6. 2022Active Learning by Feature Mixingcited by 63
  7. 2023Large-scale Dataset Pruning with Dynamic Uncertaintycited by 63
  8. 2024A Survey on Deep Active Learning: Recent Advances and New Frontierscited by 53
  9. 2021A Survey on Green Deep Learningcited by 52
  10. 2021Semi-Supervised Active Learning with Temporal Output Discrepancycited by 50
  11. 2023Dataset Quantizationcited by 47
  12. 2021Effective Evaluation of Deep Active Learning on Image Classification Taskscited by 37
38 more
  1. 2023Loss-Curvature Matching for Dataset Selection and Condensationcited by 32
  2. 2023A comprehensive survey on deep active learning in medical image analysiscited by 32
  3. 2021On the Importance of Effectively Adapting Pretrained Language Models for Active Learningcited by 31
  4. 2022GALAXY: Graph-based Active Learning at the Extremecited by 29
  5. 2025A Coreset Selection of Coreset Selection Literature: Introduction and Recent Advancescited by 28
  6. 2021Data Shapley Valuation for Efficient Batch Active Learningcited by 27
  7. 2021Unsupervised Selective Labeling for More Effective Semi-Supervised Learningcited by 27
  8. 2023Repeated Random Sampling for Minimizing the Time-to-Accuracy of Learningcited by 26
  9. 2021Black-Box Dissector: Towards Erasing-based Hard-Label Model Stealing Attackcited by 25
  10. 2022ALLSH: Active Learning Guided by Local Sensitivity and Hardnesscited by 23
  11. 2023Multi-Resolution Active Learning of Fourier Neural Operatorscited by 22
  12. 2024An Experimental Design Framework for Label-Efficient Supervised Finetuning of Large Language Modelscited by 22
  13. 2022Batch Multi-Fidelity Active Learning with Budget Constraintscited by 20
  14. 2023Streaming Active Learning with Deep Neural Networkscited by 20
  15. 2023You Only Condense Once: Two Rules for Pruning Condensed Datasetscited by 20
  16. 2021S3VAADA: Submodular Subset Selection for Virtual Adversarial Active Domain Adaptationcited by 19
  17. 2021Boosting Active Learning via Improving Test Performancecited by 19
  18. 2024SHED: Shapley-Based Automated Dataset Refinement for Instruction Fine-Tuningcited by 19
  19. 2023Algorithm Selection for Deep Active Learning with Imbalanced Datasetscited by 18
  20. 2021Label-Efficient Point Cloud Semantic Segmentation: An Active Learning Approachcited by 17
  21. 2021A Simple Baseline for Low-Budget Active Learningcited by 15
  22. 2021DeepAL: Deep Active Learning in Pythoncited by 13
  23. 2022Deep Active Learning with Noise Stabilitycited by 13
  24. 2023Re-thinking Federated Active Learning based on Inter-class Diversitycited by 13
  25. 2024Dataset Quantization with Active Learning based Adaptive Samplingcited by 13
  26. 2023Data pruning and neural scaling laws: fundamental limitations of score-based algorithmscited by 11
  27. 2021Multiple-criteria Based Active Learning with Fixed-size Determinantal Point Processescited by 9
  28. 2022Patient Aware Active Learning for Fine-Grained OCT Classificationcited by 9
  29. 2022Aries: Efficient Testing of Deep Neural Networks via Labeling-Free Accuracy Estimationcited by 9
  30. 2022Robust Active Distillationcited by 9
  31. 2023On the Limitations of Simulating Active Learningcited by 9
  32. 2023Towards Balanced Active Learning for Multimodal Classificationcited by 9
  33. 2024ELFS: Label-Free Coreset Selection with Proxy Training Dynamicscited by 9
  34. 2024Beyond Efficiency: Molecular Data Pruning for Enhanced Generalizationcited by 9
  35. 2021Stateful Detection of Model Extraction Attackscited by 8
  36. 2023Improved Algorithm for Deep Active Learning under Imbalance via Optimal Separationcited by 8
  37. 2024GDeR: Safeguarding Efficiency, Balancing, and Robustness via Prototypical Graph Pruningcited by 8
  38. 2022Deep Active Learning for Computer Vision: Past and Futurecited by 7

50 most cited of 155 All 155 (JSON)