Abstract
Training deep neural network (DNN) with noisy labels is practically challenging since inaccurate labels severely degrade the generalization ability of DNN. Previous efforts tend to handle part or full data in a unified denoising flow via identifying noisy data with a coarse small-loss criterion to mitigate the interference from noisy labels, ignoring the fact that the difficulties of noisy samples are different, thus a rigid and unified data selection pipeline cannot tackle this problem well . In this paper, we first propose a coarse-to-fine robust learning method called CREMA, to handle noisy data in a divide-and-conquer manner. In coarse-level, clean and noisy sets are firstly separated in terms of credibility in a statistical sense. Since it is practically impossible to categorize all noisy samples correctly, we further process them in a fine-grained manner via modeling the credibility of each sample. Specifically, for the clean set, we deliberately design a memory-based modulation scheme to dynamically adjust the contribution of each sample in terms of its historical credibility sequence during training, thus alleviating the effect from noisy samples incorrectly grouped into the clean set. Meanwhile, for samples categorized into the noisy set, a selective label update strategy is proposed to correct noisy labels while mitigating the problem of correction error. Extensive experiments are conducted on benchmarks of different modality, including image classification (CIFAR, Clothing1M etc.) and text recognition (IMDB), with either synthetic or natural semantic noises, demonstrating the superiority and generality of CREMA.
B. Zhang, Y.Li and Y.Tu—Contributed equally to this work.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Similar content being viewed by others
References
Arazo, E., Ortego, D., Albert, P., O’Connor, N.E., McGuinness, K.: Unsupervised label noise modeling and loss correction. In: Proc. International Conference on Machine Learning (ICML) (2019)
Arpit, D., et al.: A closer look at memorization in deep networks. In: Proc. International Conference on Machine Learning (ICML), pp. 233–242 (2017)
Berthelot, D., Carlini, N., Goodfellow, I.J., Papernot, N., Oliver, A., Raffel, C.: Mixmatch: a holistic approach to semi-supervised learning. Proc. Advances in Neural Information Processing Systems (NeurIPS) (2019)
Chang, H.S., Learned-Miller, E., McCallum, A.: Active Bias: training more accurate neural networks by emphasizing high variance samples. In: Proc. Advances in Neural Information Processing Systems (NeurIPS), pp. 1002–1012 (2017)
Dehghani, M., Severyn, A., Rothe, S., Kamps, J.: Avoiding your teacher’s mistakes: training neural networks with controlled weak supervision. (2017) arXiv preprint arXiv:1711.00313
Dehghani, M., Severyn, A., Rothe, S., Kamps, J.: Learning to learn from weak supervision by full supervision. In: Proc. Advances in Neural Information Processing Systems Workshop (NeurIPSW) (2017)
Ghosh, A., Kumar, H., Sastry, P.: Robust loss functions under label noise for deep neural networks. In: Proc. Association for the Advancement of Artificial Intelligence (AAAI) (2017)
Goldberger, J., Ben-Reuven, E.: Training deep neural-networks using a noise adaptation layer. In: Proc. International Conference on Learning Representations (ICLR) (2017)
Han, B., et al.: Masking: A new perspective of noisy supervision. In: Proc. Advances in Neural Information Processing Systems (NeurIPS), pp. 5836–5846 (2018)
Han, B., et al.: Co-teaching: robust training of deep neural networks with extremely noisy labels. In: Proc. Advances in Neural Information Processing Systems (NeurIPS), pp. 8527–8537 (2018)
Han, J., Luo, P., Wang, X.: Deep self-learning from noisy labels. In: Proc. IEEE International Conference on Computer Vision (ICCV), pp. 5138–5147 (2019)
Hendrycks, D., Mazeika, M., Wilson, D., Gimpel, K.: Using trusted data to train deep networks on labels corrupted by severe noise. In: Proc. Advances in Neural Information Processing Systems (NeurIPS), pp. 10456–10465 (2018)
Houle, M.E.: Local intrinsic dimensionality I: an extreme-value-theoretic foundation for similarity applications. In: Proc. International Conference on Similarity Search and Applications (SISAP), pp. 64–79 (2017)
Jiang, L., Zhou, Z., Leung, T., Li, L.J., Fei-Fei, L.: MentorNet: learning data-driven curriculum for very deep neural networks on corrupted labels. In: Proc. International Conference on Machine Learning (ICML) (2018)
Krizhevsky, A., Hinton, G., et al.: Learning multiple layers of features from tiny images (2009)
Li, J., Socher, R., Hoi, S.C.: Dividemix: learning with noisy labels as semi-supervised learning. (2020) arXiv preprint arXiv:2002.07394
Li, J., Wong, Y., Zhao, Q., Kankanhalli, M.: learning to learn from noisy labeled data. In: Proc. IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 5046–5054 (2019)
Li, J., Xiong, C., Hoi, S.C.: Learning from noisy data with robust representation learning. In: Proc. IEEE International Conference on Computer Vision (ICCV), pp. 9485–9494 (2021)
Li, X., Liu, T., Han, B., Niu, G., Sugiyama, M.: Provably end-to-end label-noise learning without anchor points. Proc. International Conference on Machine Learning (ICML) (2021)
Li, Y., Yang, J., Song, Y., Cao, L., Luo, J., Li, L.J.: Learning from noisy labels with distillation. In: Proc. IEEE International Conference on Computer Vision (ICCV), pp. 1910–1918 (2017)
Litany, O., Freedman, D.: Soseleto: a unified approach to transfer learning and training with noisy labels. (2018) arXiv preprint arXiv:1805.09622
Liu, T., Tao, D.: Classification with noisy labels by importance reweighting. IEEE Trans. Pattern Anal. Mach. Intell. (TPAMI) 38(3), 447–461 (2015)
Lyu, Y., Tsang, I.W.: Curriculum loss: robust learning and generalization against label corruption. In: Proc. International Conference on Learning Representations (ICLR) (2020)
Ma, X., et al.: Dimensionality-driven learning with noisy labels. In: Proc. International Conference on Machine Learning (ICML) (2018)
Maas, A., Daly, R.E., Pham, P.T., Huang, D., Ng, A.Y., Potts, C.: Learning word vectors for sentiment analysis. In: Proceedings of the 49th annual meeting of the association for computational linguistics: human language technologies, pp. 142–150 (2011)
Malach, E., Shalev-Shwartz, S.: Decoupling when to update from how to update. In: Proc. Advances in Neural Information Processing Systems (NeurIPS), pp. 960–970 (2017)
Manning, C., Schutze, H.: Foundations of statistical natural language processing. MIT press (1999)
Nguyen, D.T., Mummadi, C.K., Ngo, T.P.N., Nguyen, T.H.P., Beggel, L., Brox, T.: SELF: learning to filter noisy labels with self-Ensembling. In: Proc. International Conference on Learning Representations (ICLR) (2020)
Patrini, G., Rozza, A., Krishna Menon, A., Nock, R., Qu, L.: Making deep neural networks robust to label noise: a loss correction approach. In: Proc. IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 1944–1952 (2017)
Permuter, H., Francos, J.M., Jermyn, I.: A study of gaussian mixture models of color and texture features for image classification and segmentation. Pattern Recogn. (PR). 39, 695–706 (2006)
Pleiss, G., Zhang, T., Elenberg, E.R., Weinberger, K.Q.: Identifying mislabeled data using the area under the margin ranking. In: Proc. Advances in Neural Information Processing Systems (NeurIPS) (2020)
Reed, S., Lee, H., Anguelov, D., Szegedy, C., Erhan, D., Rabinovich, A.: Training deep neural networks on noisy labels with bootstrapping. In: Proc. International Conference on Learning Representations (ICLR) (2015)
Ren, M., Zeng, W., Yang, B., Urtasun, R.: Learning to reweight examples for robust deep learning. In: Proc. International Conference on Machine Learning (ICML) (2018)
Shu, J., et al.: Meta-Weight-Net: learning an explicit mapping for sample weighting. In: Proc. Advances in Neural Information Processing Systems (NeurIPS), pp. 1917–1928 (2019)
Song, H., Kim, M., Lee, J.G.: SELFIE: refurbishing unclean samples for robust deep learning. In: Proc. International Conference on Machine Learning (ICML), pp. 5907–5915 (2019)
Tanaka, D., Ikami, D., Yamasaki, T., Aizawa, K.: Joint optimization framework for learning with noisy labels. In: Proc. IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 5552–5560 (2018)
Veit, A., Alldrin, N., Chechik, G., Krasin, I., Gupta, A., Belongie, S.: Learning from noisy large-scale datasets with minimal supervision. In: Proc. IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2017)
Wang, R., Liu, T., Tao, D.: Multiclass learning with partially corrupted labels. IEEE Trans. Neural Networks. Learn. Syst. 29(6), 2568–2580 (2017)
Wang, X., Hua, Y., Kodirov, E., Clifton, D.A., Robertson, N.M.: Proselflc: progressive self label correction for training robust deep neural networks. In: Proc. IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 752–761 (2021)
Wang, X., Hua, Y., Kodirov, E., Robertson, N.M.: Imae for noise-robust learning: mean absolute error does not treat examples equally and gradient magnitude’s variance matters. (2019) arXiv preprint arXiv:1903.12141
Wang, X., Kodirov, E., Hua, Y., Robertson, N.M.: Derivative manipulation for general example weighting. (2019) arXiv preprint arXiv:1905.11233
Wang, Y., Ma, X., Chen, Z., Luo, Y., Yi, J., Bailey, J.: Symmetric cross entropy for robust learning with noisy labels. In: Proc. IEEE International Conference on Computer Vision (ICCV), pp. 322–330 (2019)
Wei, H., Feng, L., Chen, X., An, B.: Combating noisy labels by agreement: a joint training method with co-regularization. In: Proc. IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 13726–13735 (2020)
Xiao, T., Xia, T., Yang, Y., Huang, C., Wang, X.: Learning from massive noisy labeled data for image classification. In: Proc. IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 2691–2699 (2015)
Yao, Q., Yang, H., Han, B., Niu, G., Kwok, J.: Searching to exploit memorization effect in learning with noisy labels. In: Proc. International Conference on Machine Learning (ICML) (2020)
Yi, K., Wu, J.: Probabilistic end-to-end noise correction for learning with noisy labels. In: Proc. IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 7017–7025 (2019)
Yu, X., Han, B., Yao, J., Niu, G., Tsang, I.W., Sugiyama, M.: How does disagreement help generalization against label corruption? In: Proc. International Conference on Machine Learning (ICML) (2019)
Zhang, C., Bengio, S., Hardt, M., Recht, B., Vinyals, O.: Understanding deep learning requires rethinking generalization. In: Proc. International Conference on Learning Representations (ICLR) (2017)
Zhang, H., Cissé, M., Dauphin, Y.N., Lopez-Paz, D.: mixup: Beyond empirical risk minimization. In: Proc. International Conference on Learning Representations (ICLR) (2018)
Zhang, M., Lee, J., Agarwal, S.: Learning from noisy labels with no change to the training process. In: Proc. International Conference on Machine Learning (ICML), pp. 12468–12478 (2021)
Zhang, W., Wang, Y., Qiao, Y.: Metacleaner: Learning to hallucinate clean representations for noisy-labeled visual recognition. In: Proc. IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 7365–7374 (2019)
Zhang, Y., Zheng, S., Wu, P., Goswami, M., Chen, C.: Learning with feature-dependent label noise: a progressive approach. In: Proc. International Conference on Learning Representations (ICLR) (2021)
Zhang, Z., Sabuncu, M.: Generalized cross entropy loss for training deep neural networks with noisy labels. In: Proc. Advances in Neural Information Processing Systems (NeurIPS), pp. 8778–8788 (2018)
Zhang, Z., Zhang, H., Arik, S.O., Lee, H., Pfister, T.: Distilling effective supervision from severe label noise. In: Proc. IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 9294–9303 (2020)
Zheltonozhskii, E., Baskin, C., Mendelson, A., Bronstein, A.M., Litany, O.: Contrast to divide: self-supervised pre-training for learning with noisy labels. (2021) arXiv preprint arXiv:2103.13646
Zhou, T., Wang, S., Bilmes, J.: Robust curriculum learning: From clean label detection to noisy label self-correction. In: Proc. International Conference on Learning Representations (ICLR) (2021)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2023 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this paper
Cite this paper
Zhang, B. et al. (2023). Learning from Noisy Labels with Coarse-to-Fine Sample Credibility Modeling. In: Karlinsky, L., Michaeli, T., Nishino, K. (eds) Computer Vision – ECCV 2022 Workshops. ECCV 2022. Lecture Notes in Computer Science, vol 13802. Springer, Cham. https://doi.org/10.1007/978-3-031-25063-7_2
Download citation
DOI: https://doi.org/10.1007/978-3-031-25063-7_2
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-25062-0
Online ISBN: 978-3-031-25063-7
eBook Packages: Computer ScienceComputer Science (R0)