The DOI matches its DataCite record.
Deep Bayesian Active Learning with Image Data
Open accessAbstract
Even though active learning forms an important pillar of machine learning, deep learning tools are not prevalent within it. Deep learning poses several difficulties when used in an active learning setting. First, active learning (AL) methods generally rely on being able to learn and update models from small amounts of data. Recent advances in deep learning, on the other hand, are notorious for their dependence on large amounts of data. Second, many AL acquisition functions rely on model uncertainty, yet deep learning methods rarely represent such model uncertainty. In this paper we combine recent advances in Bayesian deep learning into the active learning framework in a practical way. We develop an active learning framework for high dimensional data, a task which has been extremely challenging so far, with very sparse existing literature. Taking advantage of specialised models such as Bayesian convolutional neural networks, we demonstrate our active learning techniques with image data, obtaining a significant improvement on existing active learning approaches. We demonstrate this on both the MNIST dataset, as well as for skin cancer diagnosis from lesion images (ISIC2016 task).
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In the citing paper
“We observe that this difference is maintained in the other cycles, including the last active learning cycle where our method outperforms the semisupervised method, multi-model methods [9, 12] and the best AL method [5] by 1 . 6% .”
From Not All Labels Are Equal: Rationalizing The Labeling Costs for Training Object Detection (2021) · cites this paper as [12]
Citations
Cites
Still being checked.
Cited by
Sorted by citations
Items 301–392 of 392
On the Robustness of Active Learning
Hahn, L., Roese-Koerner, L., Cremer, P., Zimmermann, U., Maoz, O., & Kummert, A. (2019) · EPiC series in computing · doi:10.29007/thws · Cited by 2
Loss Rescaling by Uncertainty Inference For Single-Stage Object Detection
Li, Y., Chen, X., Li, Q., & Zhang, N. (2020) · Proceedings - International Conference on Image Processing · doi:10.1109/icip40778.2020.9190845 · Cited by 2
MiikeMineStamps: A Long-Tailed Dataset of Japanese Stamps via Active Learning
Buitrago, P. A., Toropov, E., Prabha, R., Uran, J. A., & Adal, R. (2021) · Lecture notes in computer science · doi:10.1007/978-3-030-86334-0_1 · Cited by 3
Layerwise Approximate Inference for Bayesian Uncertainty Estimates on Deep Neural Networks
Zhang, N., Chen, X., & Quan, L. (2021) · International Joint Conference on Neural Networks (IJCNN) · doi:10.1109/ijcnn52387.2021.9534229 · Cited by 2
OPAD: An Optimized Policy-based Active Learning Framework for Document Content Analysis
Shekhar, S., Guda, B. P. R., Chaubey, A., Jindal, I., & Jain, A. (2022) · IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) · doi:10.1109/cvprw56347.2022.00320 · Cited by 3
Budget Active Learning for Deep Networks
Gikunda, P., & Jouandeau, N. (2020) · Advances in intelligent systems and computing · doi:10.1007/978-3-030-55180-3_36 · Cited by 2
Cluster-Based Active Learning.Preprint
Perez, F., Lebret, R., & Aberer, K. (2018) · arXiv (Cornell University) · Cited by 2
QActor: On-line Active Learning for Noisy Labeled Stream DataPreprint
Younesian, T., Zhao, Z., Ghiassi, A., Birke, R., & Chen, L. Y. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2001.10399 · Cited by 2
Information Condensing Active LearningPreprint
Jain, S., Liu, G., & Gifford, D. K. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2002.07916 · Cited by 2
Confident Coreset for Active Learning in Medical Image AnalysisPreprint
Kim, S. T., Mushtaq, F., & Navab, N. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2004.02200 · Cited by 2
MetAL: Active Semi-Supervised Learning on Graphs via Meta LearningPreprint
Madhawa, K., & Murata, T. (2020) · arXiv.org · doi:10.48550/arxiv.2007.11230 · Cited by 9
Active Learning in CNNs via Expected Improvement MaximizationPreprint
Nagpal, U. G., & Knowles, D. A. (2020) · arXiv.org · doi:10.48550/arxiv.2011.14015 · Cited by 2
Learning Novel Objects Continually Through CuriosityPreprint
Ayub, A., & Wagner, A. R. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2103.07758 · Cited by 2
BABA: Beta Approximation for Bayesian Active Learning.Preprint
Woo, J. O. (2021) · arXiv (Cornell University) · Cited by 2
A Simple Baseline for Batch Active Learning with Stochastic Acquisition Functions.Preprint
Kirsch, A., Farquhar, S., & Gal, Y. (2021) · arXiv (Cornell University) · Cited by 2
Self-citationPrioritized training on points that are learnable, worth learning, and not yet learned.Preprint
Mindermann, S., Razzak, M., Xu, W., Kirsch, A., Sharma, M., Morisot, A., Gomez, A. N., Farquhar, S., Brauner, J., & Gal, Y. (2021) · arXiv (Cornell University) · Cited by 2
Self-citationActive label cleaning: Improving dataset quality under resource constraints.Preprint
Bernhardt, M., Castro, D. C., Tanno, R., Schwaighofer, A., Tezcan, K. C., Monteiro, M., Bannur, S., Lungren, M. P., Nori, A. V., Glocker, B., et al. (2021) · arXiv (Cornell University) · Cited by 2
BioAct: Biomedical Knowledge Base Construction using Active LearningPreprint
Wright, D., Gentile, A. L., Faux, N. G., & Beck, K. L. (2022) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2022.04.14.488416 · Cited by 2
Continuous Nonlinear Adaptive Experimental Design with Gradient FlowPreprint
Jin, R., Li, Q., Mussmann, S., & Wright, S. J. (2025) · SSRN Electronic Journal · doi:10.2139/ssrn.5358687 · Cited by 2
An Effective Data Refinement Approach for Upper Gastrointestinal Anatomy Recognition
Quan, L., Li, Y., Chen, X., & Zhang, N. (2020) · Lecture Notes in Computer Science · doi:10.1007/978-3-030-59710-8_5 · Cited by 2
Understanding and Improving Neural Active Learning on Heteroskedastic Distributions
Khosla, S., Chew, K. W., Ash, J. T., Zhang, C., Kawaguchi, K., & Lamb, A. (2023) · Frontiers in artificial intelligence and applications · doi:10.3233/faia230402 · Cited by 2
Chondrogiannis, K. A., Arcieri, G., Chatzi, E., & Dertimanis, V. K. (2024) · Journal of Engineering Mechanics · doi:10.1061/jenmdt.emeng-7969 · Cited by 1
Addressing the Constraints of Active Learning on the Edge
Nueve, E., Shahkarami, S., Park, S., & Ferrier, N. (2021) · IEEE International Symposium on Parallel & Distributed Processing, Workshops and Phd Forum · doi:10.1109/ipdpsw52791.2021.00126 · Cited by 1
Weakly Supervised Learning Creates a Fusion of Modeling Cultures
Tang, C., Gan, Y., & Zheng, T. (2021) · Observational Studies · doi:10.1353/obs.2021.0011 · Cited by 2
Active Learning of Pattern Classification Based on PEDCC-Loss
Zhu, Q., Luan, J., Li, T., & Zu, X. (2021) · IEEE Access · doi:10.1109/access.2021.3123095 · Cited by 1
Interpretable Image Recognition Models for Big Data With Prototypes and Uncertainty
Wang, J. (2023) · International Journal of Information Technologies and Systems Approach · doi:10.4018/ijitsa.318122 · Cited by 1
Robust Assignment of Labels for Active Learning with Sparse and Noisy Annotations
Kałuża, D., Janusz, A., & Ślȩzak, D. (2023) · Frontiers in artificial intelligence and applications · doi:10.3233/faia230397 · Cited by 3
Diffusion-Based Sampling for Deep Active Learning
Kushnir, D., & Venturi, L. (2023) · International Conference on Sampling Theory and Applications · doi:10.1109/sampta59647.2023.10301392 · Cited by 1
Picking Groups Instead of Samples: A Close Look at Static Pool-Based Meta-Active Learning
Mas, I., Morros, J. R., & Vilaplana, V. (2019) · IEEE International Conference on Computer Vision Workshops · doi:10.1109/iccvw.2019.00171 · Cited by 1
BALD-VAE: Generative Active Learning based on the Uncertainties of Both Labeled and Unlabeled Data
Lee, S.-K., & Kim, J.-H. (2019) · International Conference on Robot Intelligence Technology and Applications · doi:10.1109/ritapp.2019.8932813 · Cited by 4
Analyzing Data Selection Techniques with Tools from the Theory of Information Losses
Foggo, B., & Yu, N. (2021) · IEEE International Conference on Big Data (BigData) · doi:10.1109/bigdata52589.2021.9671861 · Cited by 1
Collaborative Data Acquisition and Learning Support
Boiński, T. M., & Szymański, J. (2020) · Lecture notes in computer science · doi:10.1007/978-3-030-47679-3_19 · Cited by 1
VirAAL: Virtual Adversarial Active Learning for NLU
Senay, G., Idrissi, B. Y., & Haziza, M. (2021) · IEEE Spoken Language Technology Workshop (SLT) · doi:10.1109/slt48900.2021.9383576 · Cited by 2
Continual Learning of Visual Concepts for Robots through Limited Supervision
Ayub, A., & Wagner, A. R. (2021) · ACM/IEEE International Conference on Human-Robot Interaction (HRI) · doi:10.1145/3434074.3446357 · Cited by 1
Effective Version Space Reduction for Convolutional Neural Networks
Liu, J., Chiotellis, I., Triebel, R., & Cremers, D. (2021) · Lecture notes in computer science · doi:10.1007/978-3-030-67661-2_6 · Cited by 1
Krokos, V., Xuan, V. B., Bordas, S., Young, P., & Kerfriden, P. (2021) · Computational Mechanics · doi:10.1007/s00466-021-02112-3 · Cited by 73
Ragia, L., & Panagiotopoulou, A. (2021) · ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences · doi:10.5194/isprs-annals-v-3-2021-175-2021 · Cited by 1
Uncertainty Measures and Transfer Learning in Active Learning for Text Classification
Oftedal, T. O. S. (2019) · NORA - Norwegian Open Research Archives · Cited by 1
Probabilistic Semantic Segmentation Refinement by Monte Carlo Region GrowingPreprint
Dias, P. A., & Medeiros, H. P. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2005.05856 · Cited by 1
Deep Active Learning by Model InterpretabilityPreprint
Liu, Q., Liu, Z., Zhu, X., Xiu, Y.-L., & Zhu, J. (2020) · arXiv.org · doi:10.48550/arxiv.2007.12100 · Cited by 4
High Dimensional Level Set Estimation with Bayesian Neural NetworkPreprint
Ha, H., Gupta, S., Rana, S., & Venkatesh, S. (2020) · AAAI Conference on Artificial Intelligence · doi:10.48550/arxiv.2012.09973 · Cited by 11
A Simple yet Brisk and Efficient Active Learning Platform for Text ClassificationPreprint
Kanchinadam, T., You, Q., Westpfahl, K., Kim, J., Gunda, S., Seith, S., & Fung, G. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2102.00426 · Cited by 2
An efficient scheme based on graph centrality to select nodes for training for effective learningPreprint
Sandeep, C., Salim, A., Sethunadh, R., & Sumitra, S. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2104.14123 · Cited by 1
Targeted Data Acquisition for Evolving Negotiation AgentsPreprint
Kwon, M., Karamcheti, S., Cuéllar, M.-F., & Sadigh, D. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2106.07728 · Cited by 1
L2M: Practical posterior Laplace approximation with optimization-driven second moment estimationPreprint
Perone, C. S., Silveira, R. P., & Paula, T. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2107.04695 · Cited by 1
Analysis of Driving Scenario Trajectories with Active LearningPreprint
Jarl, S., Rahrovani, S., & Chehreghani, M. H. (2021) · arXiv (Cornell University) · Cited by 1
Thiagarajan, P., Khairnar, P., & Ghosh, S. (2021) · IEEE Transactions on Medical Imaging · doi:10.1109/tmi.2021.3123300 · Cited by 68
Improving Robustness and Efficiency in Active Learning with Contrastive Loss.Preprint
Krishnan, R., Ahuja, N., Sinha, A., Subedar, M. M., Tickoo, O., & Iyer, R. (2021) · arXiv (Cornell University) · Cited by 1
Deep Active Learning for Robust Biomedical SegmentationPreprint
Arikan, M., Sallo, F. B., Montesel, A., Ahmed, H., Hagag, A. M., Book, M., Faatz, H., Cicinelli, M. V., Meshkinfamfard, S., Ongun, S., et al. (2023) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2023.03.28.534521 · Cited by 1
Not All Labels Are Equal: Rationalizing The Labeling Costs for Training Object Detection
Elezi, I., Yu, Z., Anandkumar, A., Leal-Taixe, L., & Alvarez, J. M. (2021) · arXiv · Cited by 50
“Multi-model active learning networks, namely, ensemble [9] or MC-dropout [12] outperform single models at the cost of longer training and active learning time, and in the case of the ensemble has 3 times more training parameters.”
BackgroundMethodResultInfluentialIntrinsically Motivated Self-Supervised Deep Sensorimotor Learning for Grasping
Takahashi, T., Lanighan, M. W., & Grupen, R. A. (2018) · IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) · doi:10.1109/iros.2018.8593424
Learning Interpretable Models Using an Oracle.Preprint
Ghose, A., & Ravindran, B. (2019) · arXiv (Cornell University)
Bozorgtabar, B., Mahapatra, D., von Tengg‐Kobligk, H., Pöllinger, A., Ebner, L., Thiran, J.-P., & Reyes, M. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1904.10781
Variational grid setting network
Chuang, Y.-N., Huang, Z., & Tsai, Y.-L. (2017) · International Conference on Asian Language Processing · doi:10.1109/ialp.2017.8300584
Maximin Active Learning with Data-Dependent Norms
Karzand, M., & Nowak, R. D. (2019) · Allerton Conference on Communication, Control, and Computing · doi:10.1109/allerton.2019.8919686
Person Re-Identification via Active Hard Sample MiningPreprint
Xu, X., Liu, L., Liu, W., Wang, M., & Hu, R. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2004.04912
SoQal: Selective Oracle Questioning in Active LearningPreprint
Kiyasseh, D., Zhu, T., & Clifton, D. A. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2004.10468
Batch Decorrelation for Active Metric LearningPreprint
Priyadarshini, K. N., Goru, R., Chaudhuri, S., & Chaudhuri, S. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2005.10008
Variance Based Sample Weighting for Supervised Deep LearningPreprint
Novello, P., Poëtte, G., Lugato, D., & Congedo, P. M. (2021) · arXiv (Cornell University)
Mask-guided sample selection for Semi-Supervised Instance SegmentationPreprint
Bellver, M., Salvador, A., Torres, J., & Giró-i-Nieto, X. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2008.11073
Meta-Active Learning for Node Response Prediction in GraphsPreprint
Iwata, T. (2020) · arXiv.org · doi:10.48550/arxiv.2010.05387
Deep Active Graph Representation Learning.Preprint
Zhu, Y., Xu, W., Yu, F., Liu, Q., Wu, S., & Wang, L. (2020) · arXiv (Cornell University)
VAE-Info-cGANPreprint
Xiao, X., Ganguli, S., & Pandey, V. (2020) · arXiv (Cornell University)
Graph-based Deep Learning Analysis and Instance Selection
Nonaka, K., Shekkizhar, S., & Ortega, A. (2020) · IEEE International Workshop on Multimedia Signal Processing (MMSP) · doi:10.1109/mmsp48831.2020.9287121
A Taylor Based Sampling Scheme for Machine Learning in Computational PhysicsPreprint
Novello, P., Poëtte, G., Lugato, D., & Congedo, P. M. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2101.11105
Feedback Coding for Active LearningPreprint
Canal, G. H., Bloch, M., & Rozell, C. (2021) · International Conference on Artificial Intelligence and Statistics · doi:10.48550/arxiv.2103.00654 · Cited by 1
Online Learning of Objects through Curiosity-Driven Active Learning.Preprint
Ayub, A., & Wagner, A. R. (2021) · arXiv (Cornell University)
Some Nonparametric Hybrid Predictive Models : Asymptotic Properties and Applications
Chakraborty, T. (2020)
Low-Regret Active learningPreprint
Baykal, C., Liebenwein, L., Feldman, D., & Rus, D. L. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2104.02822 · Cited by 4
Mahapatra, D., Pöllinger, A., Shao, L., & Reyes Aguirre, M. A. (2021) · arXiv (Cornell University) · doi:10.48350/154465
Diversity-Aware Batch Active Learning for Dependency ParsingPreprint
Shi, T., Benton, A., Malioutov, I., & İrsoy, O. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2104.13936
Unsupervised Instance Selection with Low-Label, Supervised Learning for Outlier DetectionPreprint
Bradberry, T. J., Hase, C. H., Kent, L., & Góngora, J. A. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2104.12837
Weakly Supervised Learning Creates a Fusion of Modeling CulturesPreprint
Tang, C., Gan, Y., & Zheng, T. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2106.01485
Semi-Supervised Learning via Triplet Network Based Active Learning (Student Abstract)
Sundriyal, D., Ghosh, S., Mayank Vatsa, A., & Singh, R. (2021) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v35i18.17948
A Practical & Unified Notation for Information-Theoretic Quantities in MLPreprint
Kirsch, A., & Gal, Y. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2106.12062 · Cited by 4
Self-citationKaramcheti, S., Krishna, R., Fei-Fei, L., & Manning, C. D. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2107.02331 · Cited by 14
Reinforcement Learning Approach to Active Learning for Image ClassificationPreprint
Werner, T. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2108.05595
Face-NMS: A Core-set Selection Approach for Efficient Face RecognitionPreprint
Chen, Y., Huang, J., Zhu, J., Zhu, Z. D., Yang, T., Huang, G., & Du, D. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2109.04698 · Cited by 1
Active Learning for Argument Mining: A Practical ApproachPreprint
Solmsdorf, N., Trautmann, D., & Schütze, H. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2109.13611
Observations on K-image Expansion of Image-Mixing Augmentation for ClassificationPreprint
Jeong, J., Cha, S., Yoo, Y., Yun, S., Moon, T., & Choi, J. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2110.04248
Semi-supervised learning via triplet network based active learning
Sundriyal, D., Mayank Vatsa, A., & Singh, R. (2020)
Gitiaux, X., Khant, A., Reddy, C. K. A., Gupchup, J., & Cutler, R. (2021) · arXiv (Cornell University)
Active Learning of Neural Collision Handler for Complex 3D Mesh DeformationsPreprint
Tan, Q., Pan, Z.-R., Smith, B., Shiratori, T., & Manocha, D. (2021) · arXiv.org · doi:10.48550/arxiv.2110.07727 · Cited by 8
Dropout Prediction Variation Estimation Using Neuron Activation Strength.Preprint
Yu, H., Chen, Z., Lin, D., Shamir, G. I., & Han, J. (2021) · arXiv (Cornell University)
Bayesian Active SummarizationPreprint
Gidiotis, A., & Tsoumakas, G. (2023) · SSRN Electronic Journal · doi:10.2139/ssrn.4330828 · Cited by 4
Reducing the Long Tail Losses in Scientific Emulations with Active LearningPreprint
Lim, Y. H., & Kasim, M. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2111.08498
Active Colorization for Cartoon Line Drawings
Shuyu, C., Jiaqi, Z., Lin, G., Lai, Y., & Zhang, F. (2020) · doi:10.26686/wgtn.12585551.v1
Diverse Bayesian Active Learning with Simulated Annealing for Probabilistic SamplingPreprint
Sastry, S., Belgiu, M., & Maretto, R. V. (2022) · doi:10.36227/techrxiv.21774761
Human Interfaces with Machine Learning Recognition Systems
Clarkson, C. (2025) · doi:10.23889/suthesis.69942
(2018) · ENLIGHTEN (Jurnal Bimbingan dan Konseling Islam) · Cited by 57
Uncertainty Quantification in Deep Learning for Safer Neuroimage Enhancement
Tanno, R., Worrall, D., Kaden, E., Ghosh, A., Grussu, F., Bizzi, A., Sotiropoulos, S. N., Criminisi, A., & Alexander, D. C. (2019) · arXiv · doi:10.48550/arxiv.1907.13418 · Cited by 30
Metadata
- Authors
- Ghahramani, Z., Gal, Y., & Islam, R.
- Affiliations
- McGill University (CA)
- University of Cambridge (GB)
- Venue
- PMLR
- Publisher
- PMLR
- Published
- 2017
- Pages
- 1183-1192
- OpenAlex
- W2597787948
- Cited by
- 392
- References
- 0
- Citations by year
- 2025: 4 · 2024: 4 · 2023: 15 · 2022: 22 · 2021: 150 · 2020: 98 · 2019: 73 · 2018: 21 · 2017: 5
- FWCI
- 24.57
- Type
- conference paper
- Language
- en
- Open access
- Licence
- other-oa
- Locations
- Apollo (University of Cambridge) (submittedVersion, other-oa)
- Cambridge University Engineering Department Publications Database (submittedVersion)
- Apollo
- International Conference on Machine Learning
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Machine Learning and Algorithms
- Social Sciences › Decision Sciences › Management Science and Operations Research › Advanced Bandit Algorithms Research
- Physical Sciences › Computer Science › Artificial Intelligence › Gaussian Processes and Bayesian Inference
- Keywords
- active learning (machine learning), Bayesian deep learning, Bayesian convolutional neural network, model uncertainty, acquisition functions, image classification, high-dimensional data, skin cancer diagnosis, MNIST
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
