Learning From Noisy Labels by Regularized Estimation of Annotator Confusion
The DOI matches its Crossref record; 6 of the 6 references it cites were found in a registry or an index.
Abstract
The predictive performance of supervised learning algorithms depends on the quality of labels. In a typical label collection process, multiple annotators provide subjective noisy estimates of the ``truth" under the influence of their varying skill-levels and biases. Blindly treating these noisy labels as the ground truth limits the accuracy of learning algorithms in the presence of strong disagreement. This problem is critical for applications in domains such as medical imaging where both the annotation cost and inter-observer variability are high. In this work, we present a method for simultaneously learning the individual annotator model and the underlying true label distribution, using only noisy observations. Each annotator is modeled by a confusion matrix that is jointly estimated along with the classifier predictions. We propose to add a regularization term to the loss function that encourages convergence to the true annotator confusion matrix. We provide a theoretical argument as to how the regularization is essential to our approach both for the case of single annotator and multiple annotators. Despite the simplicity of the idea, experiments on image classification tasks with both simulated and real labels show that our method either outperforms or performs on par with the state-of-the-art methods and is capable of estimating the skills of annotators even with a single label available per image.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Gradient-based learning applied to document recognition
Lecun, Y., Bottou, L., Bengio, Y., & Haffner, P. (1998) · Proceedings of the IEEE · doi:10.1109/5.726791 · Cited by 58,805
A survey on deep learning in medical image analysis
Litjens, G., Kooi, T., Bejnordi, B. E., Setio, A. A. A., Ciompi, F., Ghafoorian, M., van der Laak, J. A. W. M., van Ginneken, B., & Sánchez, C. I. (2017) · Medical Image Analysis · doi:10.1016/j.media.2017.07.005 · Cited by 15,332
Warfield, S. K., Zou, K. H., & Wells, W. M. (2004) · IEEE Transactions on Medical Imaging · doi:10.1109/tmi.2004.828354 · Cited by 2,072
Classification in the Presence of Label Noise: A Survey
Frenay, B., & Verleysen, M. (2014) · IEEE Transactions on Neural Networks and Learning Systems · doi:10.1109/tnnls.2013.2292894 · Cited by 1,777
Maximum Likelihood Estimation of Observer Error-Rates Using the EM Algorithm
Dawid, A. P., & Skene, A. M. (1979) · Applied Statistics · doi:10.2307/2346806 · Cited by 1,692 · cited 1×
“The initial attempt was made in [10] in the early 1970s and more recently, numerous lines of research [11, 6, 12, 13, 14] proposed extensions of this work e.”
BackgroundA probabilistic u-net for segmentation of ambiguous images
Kohl, S. A. A., Romera-Paredes, B., Meyer, C., De Fauw, J., Ledsam, J. R., Maier-Hein, K. H., Eslami, S. M. A., Rezende, D. J., & Ronneberger, O. (2018) · arXiv · doi:10.48550/arxiv.1806.05034 · Cited by 439 · cited 1×
“We also assumed that there is only one ground truth for each input; this no longer holds true when the input images are truly ambiguous—recent advances in modelling multi-modality of label distributions [35, 36] potentially facilitate relaxation of such assumption.”
Background
Cited by
Learning From Noisy Labels With Deep Neural Networks: A Survey
Song, H., Kim, M., Park, D., Shin, Y., & Lee, J.-G. (2023) · IEEE Transactions on Neural Networks and Learning Systems · doi:10.1109/tnnls.2022.3152527 · Cited by 1,308
Seyyed-Kalantari, L., Zhang, H., McDermott, M. B. A., Chen, I. Y., & Ghassemi, M. (2021) · Nature Medicine · doi:10.1038/s41591-021-01595-0 · Cited by 928
“This has resulted in large annotated chest X-ray datasets 26–28 that are widely used for training deep learning models and for providing AI solutions 16,19–23 , 31 .”
BackgroundConfident Learning: Estimating Uncertainty in Dataset Labels
Northcutt, C., Jiang, L., & Chuang, I. (2021) · Journal of Artificial Intelligence Research · doi:10.1613/jair.1.12125 · Cited by 795
Confidence calibration and predictive uncertainty estimation for deep medical image segmentation
Mehrtash, A., Wells, W. M., Tempany, C. M., Abolmaesumi, P., & Kapur, T. (2020) · IEEE Transactions on Medical Imaging · doi:10.1109/tmi.2020.3006437 · Cited by 428
“addressed the correlation of ground truth quality and model uncertainty that results from rater disagreements [58], [59].”
BackgroundImage classification with deep learning in the presence of noisy labels: A survey
Algan, G., & Ulusoy, I. (2021) · Knowledge-Based Systems · doi:10.1016/j.knosys.2021.106771 · Cited by 337
Metadata
- Authors
- Tanno, R., Saeedi, A., Sankaranarayanan, S., Alexander, D. C., & Silberman, N.
- Venue
- 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)
- Publisher
- IEEE
- Published
- 2019-06
- Pages
- 11236-11245
- ISSN
- 1063-6919 (ISSN-L)
- OpenAlex
- W2979302305
- Cited by
- 237
- References
- 6
- FWCI
- 9.72
- Type
- conference paper
- Language
- en
- Open access
- closed
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
How to cite
Cite CoChat for the checks on this page, and the paper for its own findings.
- This page
- CoChat. (2026). Citation check: Learning From Noisy Labels by Regularized Estimation of Annotator Confusion. https://app.cochat.ai/papers/10.1109/cvpr.2019.01150
- The paper
- Tanno, R., Saeedi, A., Sankaranarayanan, S., Alexander, D. C., & Silberman, N. (2019). Learning From Noisy Labels by Regularized Estimation of Annotator Confusion. 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). https://doi.org/10.1109/cvpr.2019.01150
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.