Cleaning the Pool: Progressive Filtering of Unlabeled Pools in Deep Active Learning
Not on the map. Why?
Abstract (arXiv)
Existing active learning (AL) strategies capture fundamentally different notions of data value, e.g., uncertainty or representativeness. Consequently, the effectiveness of strategies can vary substantially across datasets, models, and even AL cycles. Committing to a single strategy risks suboptimal performance, as no single strategy dominates throughout the entire AL process. We introduce REFINE, an ensemble AL method that combines multiple strategies without knowing in advance which will perform best. In each AL cycle, REFINE operates in two stages: (1) Progressive filtering iteratively refines the unlabeled pool by considering an ensemble of AL strategies, retaining promising candidates capturing different notions of value. (2) Coverage-based selection then chooses a final batch from this refined pool, ensuring all previously identified notions of value are accounted for. Extensive experiments across 6 classification datasets and 3 foundation models show that REFINE consistently outperforms individual strategies and existing ensemble methods. Notably, progressive filtering serves as a powerful preprocessing step that improves the performance of any individual AL strategy applied to the refined pool, which we demonstrate on an audio spectrogram classification use case. Finally, the ensemble of REFINE can be easily extended with upcoming state-of-the-art AL strategies.
Key facts. Cited by 2 papers in the map (data through 14 Oct 2026). Builds on 23 of its 44 references that are in the map. First posted on arXiv 27 Nov 2025.
Builds on
23 of 44 references.
References · 23
- 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,744
- 2023DINOv2: Learning Robust Visual Features without Supervisioncited by 7,974
- 2025DINOv3cited by 1,414
- 2019Deep Batch Active Learning by Diverse, Uncertain Gradient Lower Boundscited by 610
- 2022Active Learning on a Budget: Opposite Strategies Suit High and Low Budgetscited by 133
- 2021Gone Fishing: Neural Active Learning with Fisher Embeddingscited by 88
- 2022Active Learning by Feature Mixingcited by 63
- 2020Towards Robust and Reproducible Active Learning Using Neural Networkscited by 54
- 2024EAT: Self-Supervised Pre-Training with Efficient Audio Transformercited by 53
- 2023Navigating the Pitfalls of Active Learning Evaluation: A Systematic Framework for Meaningful Performance Assessmentcited by 26
- 2023Algorithm Selection for Deep Active Learning with Imbalanced Datasetscited by 18
- 2023How to Select Which Active Learning Strategy is Best Suited for Your Specific Problem and Budgetcited by 17
11 more
- 2024Generalized Coverage for More Robust Low-Budget Active Learningcited by 16
- 2024Uncertainty Herding: One Active Learning Method for All Label Budgetscited by 15
- 2016Asymptotic Analysis of Objectives based on Fisher Information in Active Learningcited by 14
- 2023ActiveGLAE: A Benchmark for Deep Active Learning with Transformerscited by 14
- 2024Revisiting Active Learning in the Era of Vision Foundation Modelscited by 14
- 2025Unmute the Patch Tokens: Rethinking Probing in Multi-Label Audio Classificationcited by 7
- 2024Bridging Diversity and Uncertainty in Active learning with Self-Supervised Pre-Trainingcited by 6
- 2024Fast Fishing: Approximating BAIT for Efficient and Scalable Deep Active Image Classificationcited by 6
- 2024dopanim: A Dataset of Doppelganger Animals with Noisy Annotations from Multiple Humanscited by 3
- 2022Efficient Bayesian Updates for Deep Active Learning via Laplace Approximationscited by 2
- 2024AutoAL: Automated Active Learning with Differentiable Query Strategy Searchcited by 2
Built on by
Citing papers per year: 2026: 2.
2 papers.