# An Empirical Study on the Efficacy of Deep Active Learning for Image Classification

arXiv: https://arxiv.org/abs/2212.03088 · HTML: https://mlsanity.com/paper/2212.03088/

## Key facts

- Cited by 6 papers in the map (data through 14 Oct 2026).
- Builds on 18 of its 31 references that are in the map.
- First posted on arXiv 30 Nov 2022.

## Builds on (most cited first, 18 of 18)

- [MixMatch: A Holistic Approach to Semi-Supervised Learning](https://mlsanity.com/paper/1905.02249/) (2019)
- [Deep Bayesian Active Learning with Image Data](https://mlsanity.com/paper/1703.02910/) (2017)
- [Deep Batch Active Learning by Diverse, Uncertain Gradient Lower Bounds](https://mlsanity.com/paper/1906.03671/) (2019)
- [Pseudo-Labeling and Confirmation Bias in Deep Semi-Supervised Learning](https://mlsanity.com/paper/1908.02983/) (2019)
- [Learning Loss for Active Learning](https://mlsanity.com/paper/1905.03677/) (2019)
- [Variational Adversarial Active Learning](https://mlsanity.com/paper/1904.00370/) (2019)
- [Adversarial Active Learning for Deep Networks: a Margin Based Approach](https://mlsanity.com/paper/1802.09841/) (2018)
- [Batch Active Learning at Scale](https://mlsanity.com/paper/2107.14263/) (2021)
- [Consistency-based Semi-supervised Active Learning: Towards Minimizing Labeling Cost](https://mlsanity.com/paper/1910.07153/) (2019)
- [Deep Active Learning: Unified and Principled Method for Query and Training](https://mlsanity.com/paper/1911.09162/) (2019)
- [Diverse mini-batch Active Learning](https://mlsanity.com/paper/1901.05954/) (2019)
- [Sequential Graph Convolutional Network for Active Learning](https://mlsanity.com/paper/2006.10219/) (2020)
- [Towards Robust and Reproducible Active Learning Using Neural Networks](https://mlsanity.com/paper/2002.09564/) (2020)
- [Adversarial Sampling for Active Learning](https://mlsanity.com/paper/1808.06671/) (2018)
- [Parting with Illusions about Deep Active Learning](https://mlsanity.com/paper/1912.05361/) (2019)
- [Effective Evaluation of Deep Active Learning on Image Classification Tasks](https://mlsanity.com/paper/2106.15324/) (2021)
- [On the Marginal Benefit of Active Learning: Does Self-Supervision Eat Its Cake?](https://mlsanity.com/paper/2011.08121/) (2020)
- [Combining MixMatch and Active Learning for Better Accuracy with Fewer Labels](https://mlsanity.com/paper/1912.00594/) (2019)

## Built on by (most cited first, 6 of 6)

- [ActiveGLAE: A Benchmark for Deep Active Learning with Transformers](https://mlsanity.com/paper/2306.10087/) (2023)
- [A Cross-Domain Benchmark for Active Learning](https://mlsanity.com/paper/2408.00426/) (2024)
- [GCI-ViTAL: Gradual Confidence Improvement with Vision Transformers for Active Learning on Label Noise](https://mlsanity.com/paper/2411.05939/) (2024)
- [Towards Comparable Active Learning](https://mlsanity.com/paper/2311.18356/) (2023)
- [Balancing Accuracy, Calibration, and Efficiency in Active Learning with Vision Transformers Under Label Noise](https://mlsanity.com/paper/2505.04375/) (2025)
- [Are Candidate Models Really Needed for Active Learning?](https://mlsanity.com/paper/2605.14689/) (2026)

## Abstract (arXiv)

Deep Active Learning (DAL) has been advocated as a promising method to reduce labeling costs in supervised learning. However, existing evaluations of DAL methods are based on different settings, and their results are controversial. To tackle this issue, this paper comprehensively evaluates 19 existing DAL methods in a uniform setting, including traditional fully-underlinesupervised underlineactive underlinelearning (SAL) strategies and emerging underlinesemi-underlinesupervised underlineactive underlinelearning (SSAL) techniques. We have several non-trivial findings. First, most SAL methods cannot achieve higher accuracy than random selection. Second, semi-supervised training brings significant performance improvement compared to pure SAL methods. Third, performing data selection in the SSAL setting can achieve a significant and consistent performance improvement, especially with abundant unlabeled data. Our findings produce the following guidance for practitioners: one should (i) apply SSAL early and (ii) collect more unlabeled data whenever possible, for better model performance.