CoChat
Sign Up
Cited by 196Adversarial Robustness in Machine LearningMachine Learning and Data ClassificationAdvanced Neural Network Applications

Human Uncertainty Makes Classification More Robust

Open access
Peterson, Joshua, Battleday, Ruairidh, Griffiths, Thomas, Russakovsky, Olga·2019·2019 IEEE/CVF International Conference on Computer Vision (ICCV)
DOI 10.1109/iccv.2019.00971
Open this paper in CoChat — free accountOpen PDF

Abstract

The classification performance of deep neural networks has begun to asymptote at near-perfect levels. However, their ability to generalize outside the training set and their robustness to adversarial attacks have not. In this paper, we make progress on this problem by training with full label distributions that reflect human perceptual uncertainty. We first present a new benchmark dataset which we call CIFAR10H, containing a full distribution of human labels for each image of the CIFAR10 test set. We then show that, while contemporary classifiers fail to exhibit human-like uncertainty on their own, explicit training on our dataset closes this gap, supports improved generalization to increasingly out-of-training-distribution test datasets, and confers robustness to adversarial attacks.

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1109/iccv.2019.00971

Cites

7 of 7 references verified0 citation errors · 0 not checkable

Statistics

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

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?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Cites

Flagged first, then by citations

Research works printed but not linked: 45

  1. Jonathan Bragg, Mausam, and Daniel S. Weld. Crowd- sourcing multi-label classification for taxonomy creation. In Conference on Human Computation and Crowdsourcing (HCOMP), 2013.
  2. Steve Branson, Catherine Wah, Florian Schroff, Boris Babenko, Peter Welinder, Pietro Perona, and Serge Belongie. Visual recognition with humans in the loop. In European Conference on Computer Vision (ECCV), 2010.
  3. Michael Buhrmester, Tracy Kwang, and Samuel D Gosling. Amazon's mechanical turk: A new source of inexpensive, yet high-quality, data? Perspectives on Psychological Science, 6(1):3-5, 2011.
  4. L. B. Chilton, G. Little, D. Edge, D. S. Weld, and J. A. Lan- day. Cascade: Crowdsourcing taxonomy creation. In Confer- ence on Human Factors in Computing Systems (CHI), 2013.
  5. Luke Nicholas Darlow, Elliot J. Crowley, Antreas Antoniou, and Amos J. Storkey. CINIC-10 is not imagenet or CIFAR- 10. arXiv preprint arXiv:1810.03505, 2018. · arXiv
  6. Jia Deng, Nan Ding, Yangqing Jia, Andrea Frome, Kevin Murphy, Samy Bengio, Yuan Li, Hartmut Neven, and Hartwig Adam. Large-scale object classification using la- bel relation graphs. In European Conference on Computer Vision (ECCV), 2014.
  7. R. Fergus, H. Bernal, Y. Weiss, and A. Torralba. Semantic label sharing for learning with many categories. In European Conference on Computer Vision (ECCV), 2010.
  8. A. Frome, G.S. Corrado, J. Shlens, S. Bengio, J. Dean, and T. Mikolov. Devise: A deep visual-semantic embedding model. In Advances in Neural Information Processing Sys- tems (NeurIPS), 2013.
  9. Xavier Gastaldi. Shake-shake regularization. arXiv preprint arXiv:1705.07485, 2017. · arXiv
  10. Aritra Ghosh, Himanshu Kumar, and PS Sastry. Robust loss functions under label noise for deep neural networks. In Con- ference on Artificial Intelligence (AAAI), 2017.
  11. G. Griffin and P. Perona. Learning and using taxonomies for fast visual categorization. In Computer Vision and Pattern Recognition (CVPR), 2008.
  12. Dongyoon Han, Jiwhan Kim, and Junmo Kim. Deep pyra- midal residual networks. In Conference on Computer Vision and Pattern Recognition (CVPR), 2017.
  13. Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. Identity mappings in deep residual networks. In European Conference on Computer Vision (ECCV), 2016.
  14. Joel Hestness, Sharan Narang, Newsha Ardalani, Gregory F. Diamos, Heewoo Jun, Hassan Kianinejad, Md. Mostofa Ali Patwary, Yang Yang, and Yanqi Zhou. Deep learning scaling is predictable, empirically. arXiv preprint arXiv:1712.00409, 2017. · arXiv
  15. Geoffrey E. Hinton, Oriol Vinyals, and Jeffrey Dean. Dis- tilling the knowledge in a neural network. arXiv preprint arXiv:1503.02531, 2015. · arXiv
  16. Gao Huang, Zhuang Liu, Laurens Van Der Maaten, and Kil- ian Q Weinberger. Densely connected convolutional net- works. In Computer Vision and Pattern Recognition (CVPR), 2017.
  17. Yanping Huang, Yonglong Cheng, Dehao Chen, Hy- oukJoong Lee, Jiquan Ngiam, Quoc V Le, and Zhifeng Chen. Gpipe: Efficient training of giant neural networks us- ing pipeline parallelism. arXiv preprint arXiv:1811.06965, 2018. · arXiv
  18. S.J. Hwang, F. Sha, and K. Grauman. Sharing features be- tween objects and their attributes. In Computer Vision and Pattern Recognition (CVPR), 2011.
  19. Y. Jia, J.T. Abbott, J. Austerweil, T. Griffiths, and T. Dar- rell. Visual concept learning: Combining machine vision and bayesian generalization on concept hierarchies. In Advances in Neural Information Processing Systems (NeurIPS), 2013.
  20. Diederik P. Kingma and Jimmy Ba. Adam: A method for stochastic optimization. In International Conference on Learning Representations (ICLR), 2014.
  21. Adriana Kovashka, Olga Russakovsky, Li Fei-Fei, and Kristen Grauman. Crowdsourcing in Computer Vision. Foundation and Trends in Computer Graphics and Vision, 10(3):177-243, 2016.
  22. Ranjay A. Krishna, Kenji Hata, Stephanie Chen, Joshua Kravitz, David A. Shamma, Li Fei-Fei, and Michael S. Bern- stein. Embracing error to enable rapid crowdsourcing. In Conference on Human Factors in Computing Systems (CHI), 2016.
  23. Alexey Kurakin, Ian Goodfellow, and Samy Bengio. Ad- versarial examples in the physical world. In International Conference on Learning Representations (ICLR), 2016.
  24. George Lakoff. Women, fire, and dangerous things. Univer- sity of Chicago press, 2008.
  25. C.H. Lampert. Maximum margin multi-label structured pre- diction. In Advances in Neural Information Processing Sys- tems (NeurIPS), 2011.
  26. Tsung-Yi Lin, Michael Maire, Serge Belongie, James Hays, Pietro Perona, Deva Ramanan, Piotr Dollár, and C Lawrence Zitnick. Microsoft coco: Common objects in context. In European Conference on Computer Vision (ECCV), 2014.
  27. M. Marszalek and C. Schmid. Semantic hierarchies for vi- sual object recognition. In Computer Vision and Pattern Recognition (CVPR), 2007.
  28. An Thanh Nguyen, Byron C Wallace, and Matthew Lease. Combining crowd and expert labels using decision theoretic active learning. In Conference on Human Computation and Crowdsourcing (HCOMP), 2015.
  29. Quang Nguyen, Hamed Valizadegan, and Milos Hauskrecht. Learning classification models with soft-label information. Journal of the American Medical Informatics Association, 21(3):501-508, 2014.
  30. Nicolas Papernot, Patrick McDaniel, Xi Wu, Somesh Jha, and Ananthram Swami. Distillation as a defense to adver- sarial perturbations against deep neural networks. In IEEE Symposium on Security and Privacy (SP), pages 582-597. IEEE, 2016.
  31. Adam Paszke, Sam Gross, Soumith Chintala, Gregory Chanan, Edward Yang, Zachary DeVito, Zeming Lin, Al- ban Desmaison, Luca Antiga, and Adam Lerer. Automatic differentiation in PyTorch. In NeurIPS Autodiff Workshop, 2017.
  32. Benjamin Recht, Rebecca Roelofs, Ludwig Schmidt, and Vaishaal Shankar. Do cifar-10 classifiers generalize to cifar- 10? arXiv preprint arXiv:1806.00451, 2018. · arXiv
  33. David Rolnick, Andreas Veit, Serge J. Belongie, and Nir Shavit. Deep learning is robust to massive label noise. arXiv preprint arXiv:1705.10694, 2017. · arXiv
  34. Olga Russakovsky, Jia Deng, Hao Su, Jonathan Krause, San- jeev Satheesh, Sean Ma, Zhiheng Huang, Andrej Karpathy, Aditya Khosla, Michael Bernstein, et al. Imagenet large scale visual recognition challenge. International Journal of Computer Vision, 115(3):211-252, 2015.
  35. Karen Simonyan and Andrew Zisserman. Very deep con- volutional networks for large-scale image recognition. In In- ternational Conference on Learning Representations (ICLR), 2014.
  36. Christian Szegedy, Vincent Vanhoucke, Sergey Ioffe, Jon Shlens, and Zbigniew Wojna. Rethinking the inception ar- chitecture for computer vision. In Computer Vision and Pat- tern Recognition (CVPR), 2016.
  37. Christian Szegedy, Wojciech Zaremba, Ilya Sutskever, Joan Bruna, Dumitru Erhan, Ian Goodfellow, and Rob Fergus. In- triguing properties of neural networks. In International Con- ference on Learning Representations (ICLR), 2013.
  38. Antonio Torralba, Rob Fergus, and William T Freeman. 80 million tiny images: A large data set for nonparametric object and scene recognition. IEEE Transactions on Pat- tern Analysis and Machine Intelligence, 30(11):1958-1970, 2008.
  39. Antonio Torralba, Kevin P Murphy, and William T Freeman. Sharing visual features for multiclass and multiview object detection. IEEE Transactions on Pattern Analysis and Ma- chine Intelligence, 29(5):854-869, 2007.
  40. Arash Vahdat. Toward robustness against label noise in train- ing deep discriminative neural networks. In Advances in Neural Information Processing Systems (NeurIPS). 2017.
  41. Saining Xie, Ross B Girshick, Piotr Dollár, Zhuowen Tu, and Kaiming He. Aggregated residual transformations for deep neural networks. In Computer Vision and Pattern Recogni- tion (CVPR), 2017.
  42. Sergey Zagoruyko and Nikos Komodakis. Wide residual net- works. In British Machine Vision Conference (BMVC), 2016.
  43. Hongyi Zhang, Moustapha Cisse, Yann N Dauphin, and David Lopez-Paz. mixup: Beyond empirical risk minimiza- tion. In International Conference on Learning Representa- tions (ICLR), 2017.
  44. Bolei Zhou, Agata Lapedriza, Aditya Khosla, Aude Oliva, and Antonio Torralba. Places: A 10 million image database for scene recognition. IEEE Transactions on Pattern Analy- sis and Machine Intelligence, 2017.
  45. A. Zweig and D. Weinshall. Exploiting object hierarchy: combining models from different category levels. In Inter- national Conference on Computer Vision (ICCV), 2007.

Cited by

Sorted by citations

Authors
Peterson, J., Battleday, R., Griffiths, T., & Russakovsky, O.
Affiliations
Princeton University (US)
ORCID
Venue
2019 IEEE/CVF International Conference on Computer Vision (ICCV)
Publisher
IEEE
Published
2019-10
Pages
9616-9625
DOI
10.1109/iccv.2019.00971
OpenAlex
W2969769592
Cited by
196
References
7
Citations by year
2026: 15 · 2025: 34 · 2024: 25 · 2023: 40 · 2022: 28 · 2021: 36 · 2020: 17 · 2019: 1
FWCI
7.84
Type
conference paper
Language
en
Open access
Licence
Locations
  • IEEE/CVF International Conference on Computer Vision (ICCV) (publishedVersion)
  • arXiv (Cornell University) (submittedVersion)
  • arXiv (Cornell University)
Funders
National Science Foundation (1718550)
Topics
  • Physical Sciences › Computer Science › Artificial Intelligence › Adversarial Robustness in Machine Learning
  • Physical Sciences › Computer Science › Artificial Intelligence › Machine Learning and Data Classification
  • Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Neural Network Applications
Keywords
human uncertainty, label distribution, out-of-distribution generalization, adversarial robustness, uncertainty estimation, deep neural network, benchmark dataset, perceptual uncertainty

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