Pith. sign in

REVIEW 4 major objections 5 minor 64 references

Are Synthetic Corruptions A Reliable Proxy For Real-World Corruptions?

T0 review · 4 major / 5 minor · reviewed 2026-08-15 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read The paper claims that, for relative robustness evaluation, synthetic corruptions rank semantic segmentation models nearly the same as real-world adverse conditions.

desk verdict The benchmark is genuinely useful, but the paper's central correlation claim lacks the controls needed to separate corruption-specific signal from generic model capacity. read the letter →

arxiv 2505.04835 v1 pith:EDBTF2XL submitted 2025-05-07 cs.CV

Shashank Agnihotri , David Schader , Nico Sharei , Mehmet Ege Kaçar , Margret Keuper This is my paper · ORCID
classification cs.CV
keywords syntheticcorruptionsreal-worldrobustnessevaluationsemanticsegmentationdistributionshiftACDCCommonPearsoncorrelation
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper asks whether synthetic image corruptions, which are cheap to generate and widely used, can take the place of real-world adverse-condition data when evaluating how robust a semantic segmentation model is. It benchmarks a large set of models trained on the Cityscapes dataset against both 2D Common Corruptions and the ACDC real-world adverse-conditions dataset, and compares the two orderings of models. The central finding is a Pearson correlation of about 0.795 between mean performance on synthetic corruptions at severity 3 and performance on real-world corruptions, with an even higher correlation of 0.828 for worst-case synthetic performance. The paper therefore concludes that, for relative analysis, ranking or selecting among models, synthetic corruptions are a reliable proxy for real-world robustness, even though individual corruption types vary in how well they align.

What carries the argument

The central object is a correlation benchmark pairing two test suites: synthetic 2D Common Corruptions, namely the 15 perturbation types applied at severity 3 to Cityscapes validation images, and ACDC, a dataset of images captured in real adverse conditions. The load-bearing comparison is the Pearson correlation between models' mean mIoU on the two suites, complemented by GAM3, the worst mIoU across all synthetic corruptions at severity 3, which captures the model's weakest synthetic condition. This design lets the paper ask both aggregate questions about whether the model ranking transfers and corruption-specific questions about which synthetic perturbations predict which real conditions.

What would settle it

Compute the partial correlation between mean synthetic-corruption mIoU and ACDC mIoU after controlling for clean Cityscapes mIoU or model parameter count; if the partial correlation falls to near zero, the apparent proxy signal is mostly general model quality rather than corruption-specific robustness. Alternatively, test a set of models matched for clean accuracy but known to differ in corruption robustness: if their synthetic and real rankings diverge, the proxy claim fails.

Watch

Extended reading notes

Core claim

The paper's claim is that, for relative analysis, synthetic corruptions do serve as a reliable proxy for real-world corruptions in semantic segmentation. The evidence is a large benchmarking study: models trained on Cityscapes and spanning many architectures and backbones are evaluated on the 15 synthetic 2D Common Corruptions at severity 3 and on the real-world ACDC dataset under night, fog, rain, and snow. Mean mIoU across synthetic corruptions correlates with ACDC mIoU at a Pearson correlation of about 0.795, and the worst-case synthetic measure GAM3 correlates at 0.828, so the ordering of models by synthetic robustness closely matches their ordering by real-world robustness. Corruption-specific analysis finds that synthetic snow aligns strongly with real snow at 0.867, frost correlates moderately with real fog at 0.680, while synthetic brightness versus real night at 0.270 and synthetic fog versus real fog at 0.349 align weakly. The paper reads this as evidence that synthetic corruptions work as a proxy at the level of aggregate model ranking, while specific corruption types need refinement.

Load-bearing premise

The analysis assumes that ACDC stands in for real-world corruptions generally and that the correlation across models reflects shared corruption-robustness rather than simply better models scoring higher on every test set.

Editorial extensions

If this is right

  • If the correlation holds, robustness evaluation for semantic segmentation can be run largely on synthetic corruptions, avoiding the cost of collecting real-world adverse-condition images.
  • The worst-case synthetic measure GAM3 predicts real-world performance at least as well as mean synthetic performance, so a robustness test built around the model's weakest corruption is a sensible design choice.
  • Synthetic snow is a trustworthy stand-in for real snow, while synthetic brightness and fog are not; benchmarks targeting nighttime or foggy driving should supplement synthetic tests with real data.
  • The released benchmarking numbers for Cityscapes, ADE20K, and PASCAL VOC2012 give practitioners reference points for comparing future models on synthetic corruptions.
  • The claim is explicitly about relative analysis; absolute mIoU on synthetic data should not be read as the expected absolute mIoU in the field.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Beyond the paper: the aggregate correlation might be partly explained by general model quality, since better models tend to score higher on every test set; a partial-correlation control for clean Cityscapes mIoU or parameter count would show whether the synthetic-to-real link survives once general quality is removed.
  • Beyond the paper: the weak fog correlation suggests a concrete improvement path, namely physically based fog simulation or depth-aware rendering, rather than the current 2D synthetic fog, before synthetic fog can rank models for real fog.
  • Beyond the paper: a practitioner could use synthetic corruptions to pre-screen a large model zoo and then validate only the top candidates on real-world data; the paper's relative-analysis framing supports this two-stage workflow, though it tests no such protocol.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 5 minor

Summary. The paper benchmarks a large collection of semantic segmentation models trained on Cityscapes by evaluating them on the real-world ACDC dataset and on 2D Common Corruptions applied to Cityscapes at severity 3. It reports Pearson correlations between mean mIoU on synthetic corruptions and mIoU on ACDC, finding values around 0.8, and interprets this as evidence that synthetic corruptions are a reliable proxy for real-world robustness for relative model comparisons. The paper also reports corruption-specific correlations (snow 0.867, brightness 0.270, fog 0.349, frost vs. fog 0.680) and provides benchmarking results on ADE20K and PASCAL VOC. The central conclusion is that synthetic corruptions support robustness evaluation, while individual corruption types are not universally reliable proxies.

Significance. If the reported correlation is genuinely corruption-specific rather than an artifact of general model quality, the finding is practically valuable: it would justify using cheap synthetic corruptions for model selection and robustness screening in semantic segmentation. The benchmark itself, spanning many architectures and backbones, is a useful community resource. However, the key statistical claim currently lacks controls for clean performance and model capacity, does not report confidence intervals or the number of models used, and contains an unresolved discrepancy in the headline correlation. The significance of the paper therefore hinges on additional analysis that the manuscript does not provide.

major comments (4)
  1. [Sec. 4.1, Figs. 2 and 3] The same comparison, mean 2D Common Corruption mIoU versus ACDC mIoU, is reported with Pearson correlation 0.795 in Fig. 2 and 0.759 in Fig. 3. Because this is the headline evidence for the paper's central claim, the discrepancy must be resolved, and the exact number and identity of models used in each scatter plot should be stated. Without a consistent model set, the reproducibility of the main correlation is in question.
  2. [Sec. 4.1 and Appendix Fig. 8] The r approximately 0.8 correlation is not shown to be corruption-specific. Appendix Fig. 8 shows that mIoU on most individual corruptions rises with parameter count, and its caption acknowledges that the same architecture/backbone pair appears twice due to two different crop sizes; these duplicated, non-independent points may inflate the reported correlations. The paper should control for general model quality, for example by reporting partial correlations with clean Cityscapes validation mIoU and log parameter count, or by comparing the synthetic-vs-ACDC correlation with the clean-vs-ACDC correlation. Without such a control, the corruption-specific conclusions of Sec. 4.2, such as snow at 0.867 versus brightness at 0.270, may reflect general capacity rather than a shared corruption-specific signal.
  3. [Sec. 4.2] The corruption-specific Pearson correlations are reported without confidence intervals, p-values, or the number of models used in each analysis. With an unreported sample size and multiple comparisons across corruption types, the qualitative claims of 'very strong alignment' for snow and 'weak alignment' for brightness and fog are not statistically supported. The authors should provide uncertainty quantification or a rank-based statistic, and should also justify the use of Pearson r if the scatter plots reveal nonlinear monotone relationships.
  4. [Title, abstract, and conclusion] The paper's central claim is stated for 'real-world corruptions' in general, but the only real-world dataset used is ACDC, which the related-work section itself describes as covering only a limited set of conditions 'to a certain extent'. The title and abstract should be qualified to 'adverse weather and lighting conditions' or the paper should include additional real-world datasets before making the broader generalization.
minor comments (5)
  1. [Sec. 4.1] The sentence 'we prove this assumption' is too strong for an empirical correlation analysis; suggest using 'provide evidence for' or 'test'.
  2. [Sec. 4.1 and Table 1] The main text should clarify that Figs. 2 and 3 use only the Cityscapes-trained subset of Table 1, and should report the resulting sample size n in each panel.
  3. [Appendix Fig. 8 caption] The duplicate architecture/backbone pairs caused by two crop sizes are disclosed in the appendix, but this non-independence should also be mentioned in the main-text discussion of Figs. 2 and 3, and markers should distinguish the two crop sizes.
  4. [Appendix A] Appendix A reports correlation ranges such as 0.782-0.858 without stating what the range spans; please specify the exact comparisons and model sets used to produce each number.
  5. [Appendix F.1] The limitations paragraph acknowledges the lack of severity-level analysis, but it does not address the capacity confound identified above; a brief sentence acknowledging this would help calibrate the reader's expectations.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the central correlation claim is an empirical benchmark result resting on external datasets, not on fitted inputs or self-citation.

full rationale

The paper's central claim is that mean mIoU on 2D Common Corruptions (severity 3) correlates strongly with mIoU on the external ACDC dataset across a large set of Cityscapes-trained semantic segmentation models. This is an empirical comparison of two independently defined benchmark evaluations: ACDC is an external real-world dataset, and the synthetic corruptions are standard Hendrycks-Dietterich corruptions. No parameter is fitted to ACDC, and the GAM measure is simply the minimum mIoU across synthetic corruptions, a definition rather than a fitted quantity. The paper's own self-citations appear in the related-work and method descriptions, but none of them supplies the load-bearing evidence for the correlation; the evidence is the reported benchmark itself. Concerns such as whether the correlation is driven by generic model capacity or whether ACDC exhaustively represents real-world corruptions are substantive external-validity and statistical-control issues, not circularity: the derivation does not assume its conclusion. Accordingly, no circular step can be exhibited, and the honest finding is no significant circularity.

Assumptions & free parameters 0 free parameters · 4 assumptions · 0 invented entities

The paper introduces no new physical entities and fits no free parameters. Its load-bearing assumptions are domain choices: ACDC as the real-world reference, severity=3 for synthetic corruptions, representativeness of the model sample, and the use of Pearson correlation as the proxy criterion.

assumptions (4)
  • domain assumption ACDC reliably represents the class of real-world corruptions relevant to semantic segmentation in driving scenes.
    The paper uses ACDC [50] as the only real-world corruptions benchmark and generalizes from it to 'real-world corruptions'. ACDC contains only night, fog, rain, and snow conditions from the same geographic region as Cityscapes.
  • domain assumption 2D Common Corruptions at severity level 3 are a meaningful synthetic proxy for real-world distribution shifts.
    The paper adopts severity=3 based on [31,46] without re-validating severity levels for the proxy question. Appendix F.1 notes that correlation with different severity levels is left for future work.
  • domain assumption The set of evaluated model architectures and backbones is representative of semantic segmentation methods.
    The correlation is computed over a finite set of models listed in Table 1, all trained on Cityscapes for the main analysis. No evidence is given that this sample spans the full model space.
  • domain assumption Pearson correlation across models is an appropriate measure of 'reliability as a proxy'.
    The paper equates high Pearson correlation with proxy reliability, without considering rank-based metrics, prediction intervals, or per-model absolute error.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Are Synthetic Corruptions A Reliable Proxy For Real-World Corruptions?." pith.science (2026). https://pith.science/paper/EDBTF2XL

@misc{pith2026250504835,
  author       = {Pith},
  title        = {Pith review of: Are Synthetic Corruptions A Reliable Proxy For Real-World Corruptions?},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/EDBTF2XL}},
  note         = {Machine review of arXiv:2505.04835}
}
read the original abstract

Deep learning (DL) models are widely used in real-world applications but remain vulnerable to distribution shifts, especially due to weather and lighting changes. Collecting diverse real-world data for testing the robustness of DL models is resource-intensive, making synthetic corruptions an attractive alternative for robustness testing. However, are synthetic corruptions a reliable proxy for real-world corruptions? To answer this, we conduct the largest benchmarking study on semantic segmentation models, comparing performance on real-world corruptions and synthetic corruptions datasets. Our results reveal a strong correlation in mean performance, supporting the use of synthetic corruptions for robustness evaluation. We further analyze corruption-specific correlations, providing key insights to understand when synthetic corruptions succeed in representing real-world corruptions. Open-source Code: https://github.com/shashankskagnihotri/benchmarking_robustness/tree/segmentation_david/semantic_segmentation

Figures

Figures reproduced from arXiv: 2505.04835 by the authors.

Figure 1
Figure 1. Comparing images with weather corruptions captured in the wild (ACDC [ [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. To empirically determine if synthetic common corruptions such as those proposed by [ [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Correlation between model performance (legend as in Fig. [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (6 more)
Figure 4
Figure 4. Figure 4: Comparison of mean accuracy across synthetic (2D Common Corruptions) and real-world (ACDC) corruptions. The top plot [PITH_FULL_IMAGE:figures/full_fig_p009_4.png]
Figure 5
Figure 5. Figure 5: Illustrating changes in prediction due to different 2D Common Corruptions on a randomly chosen input image from the [PITH_FULL_IMAGE:figures/full_fig_p012_5.png]
Figure 6
Figure 6. Figure 6: An extension to Figure [PITH_FULL_IMAGE:figures/full_fig_p013_6.png]
Figure 7
Figure 7. Figure 7: Dataset used: PASCAL VOC2012. The correlation in the performance of semantic segmentation methods against different 2D Common Corruptions. The respective axis shows the name of the common corruption used. Colors are used to show different architectures and marker style…
Figure 8
Figure 8. Figure 8: Dataset used: Cityscapes. The correlation in the performance of semantic segmentation methods against different 2D Common Corruptions. The respective axis shows the name of the common corruption used. Colors are used to show different architectures and marker styles ar…
Figure 9
Figure 9. Figure 9: Dataset used: ADE20K. The correlation in the performance of semantic segmentation methods against different 2D Common Corruptions. The respective axis shows the name of the common corruption used. Colors are used to show different architectures and marker styles are us…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

64 extracted references · 46 canonical work pages

  1. [1]

    On the unreasonable vulnerability of transformers for image restoration-and an easy fix

    Shashank Agnihotri, Kanchana Vaishnavi Gandikota, Julia Grabinski, Paramanand Chandramouli, and Margret Keuper. On the unreasonable vulnerability of transformers for image restoration-and an easy fix. In Proc. of the IEEE/CVF In- ternational Conference on Computer Vision (ICCV) , pages 3707–3717, 2023. 2

    work page 2023

  2. [2]

    Improving stability during upsampling – on the importance of spatial context, 2023

    Shashank Agnihotri, Julia Grabinski, and Margret Keuper. Improving stability during upsampling – on the importance of spatial context, 2023. 2

    work page 2023

  3. [3]

    Beware of aliases–signal preservation is crucial for robust image restoration

    Shashank Agnihotri, Julia Grabinski, Janis Keuper, and Margret Keuper. Beware of aliases–signal preservation is crucial for robust image restoration. arXiv preprint arXiv:2406.07435, 2024. 2

    arXiv 2024

  4. [4]

    CosPGD: an efficient white-box adversarial attack for pixel- wise prediction tasks

    Shashank Agnihotri, Steffen Jung, and Margret Keuper. CosPGD: an efficient white-box adversarial attack for pixel- wise prediction tasks. In Proc. International Conference on Machine Learning (ICML), 2024. 2, 3, 9

    work page 2024

  5. [5]

    Roll the dice: Monte carlo downsampling as a low-cost adversarial defence, 2024

    Shashank Agnihotri, Shashank Priyadarshi, Hendrik Som- merhoff, Julia Grabinski, Andreas Kolb, and Margret Keu- per. Roll the dice: Monte carlo downsampling as a low-cost adversarial defence, 2024. 2

    work page 2024

  6. [6]

    DispBench

    Shashank Agnihotri, Amaan Ansari, Annika Dackermann, Fabian R ¨osch, and Margret Keuper. DispBench. In CVPR Workshop On Synthetic Data for Computer Vision, 2025. 2

    work page 2025

  7. [7]

    FlowBench: A Robustness Benchmark for Optical Flow Estimation, 2025

    Shashank Agnihotri, Julian Yuya Caspary, Luca Schwarz, Xinyan Gao, Jenny Schmalfuss, Andres Bruhn, and Margret Keuper. FlowBench: A Robustness Benchmark for Optical Flow Estimation, 2025. 8

    work page 2025

  8. [8]

    On the robustness of semantic segmentation models to adversarial attacks

    Anurag Arnab, Ondrej Miksik, and Philip HS Torr. On the robustness of semantic segmentation models to adversarial attacks. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 888–897, 2018. 3

    work page 2018

Show all 64 references
  1. [9]

    Beit: Bert pre-training of image transformers

    Hangbo Bao, Li Dong, Songhao Piao, and Furu Wei. Beit: Bert pre-training of image transformers. arXiv preprint arXiv:2106.08254, 2021. 11

    2021 arXiv

  2. [10]

    Segmentmeifyoucan: A benchmark for anomaly segmentation

    Robin Chan, Krzysztof Lis, Svenja Uhlemeyer, Hermann Blum, Sina Honari, Roland Siegwart, Pascal Fua, Mathieu Salzmann, and Matthias Rottmann. Segmentmeifyoucan: A benchmark for anomaly segmentation. In Proceedings of the Neural Information Processing Systems Track on Datasets ...

    2021

  3. [11]

    Rethinking atrous convolution for semantic image segmentation, 2017

    Liang-Chieh Chen, George Papandreou, Florian Schroff, and Hartwig Adam. Rethinking atrous convolution for semantic image segmentation, 2017. 11

    2017

  4. [12]

    Encoder-decoder with atrous separable convolution for semantic image segmentation

    Liang-Chieh Chen, Yukun Zhu, George Papandreou, Florian Schroff, and Hartwig Adam. Encoder-decoder with atrous separable convolution for semantic image segmentation. In Proceedings of the European conference on computer vision (ECCV), pages 801–818, 2018. 10, 11

    2018

  5. [13]

    Schwing, Alexan- der Kirillov, and Rohit Girdhar

    Bowen Cheng, Ishan Misra, Alexander G. Schwing, Alexan- der Kirillov, and Rohit Girdhar. Masked-attention mask transformer for universal image segmentation. In CVPR,

  6. [14]

    Xception: Deep learning with depthwise separable convolutions

    Franc ¸ois Chollet. Xception: Deep learning with depthwise separable convolutions. In Proceedings of the IEEE con- ference on computer vision and pattern recognition , pages 1251–1258, 2017. 10

    2017

  7. [15]

    The cityscapes dataset for semantic urban scene understanding

    Marius Cordts, Mohamed Omran, Sebastian Ramos, Timo Rehfeld, Markus Enzweiler, Rodrigo Benenson, Uwe Franke, Stefan Roth, and Bernt Schiele. The cityscapes dataset for semantic urban scene understanding. In Proceed- ings of the IEEE conference on computer vision and pattern re...

    2016

  8. [16]

    RobustBench: a standardized adversarial robustness benchmark

    Francesco Croce, Maksym Andriushchenko, Vikash Se- hwag, Edoardo Debenedetti, Nicolas Flammarion, Mung Chiang, Prateek Mittal, and Matthias Hein. RobustBench: a standardized adversarial robustness benchmark. InAdvances in Neural Information Processing Systems (NeurIPS), 2021. 3

    2021

  9. [17]

    2 net for generalized zero-label semantic seg- mentation

    Anurag Das, Yongqin Xian, Yang He, Bernt Schiele, and Zeynep Akata. 2 net for generalized zero-label semantic seg- mentation. In DAGM German Conference on Pattern Recog- nition, pages 235–249. Springer, 2021. 2

    2021

  10. [18]

    Weakly-supervised domain adaptive semantic segmentation with prototypical contrastive learning

    Anurag Das, Yongqin Xian, Dengxin Dai, and Bernt Schiele. Weakly-supervised domain adaptive semantic segmentation with prototypical contrastive learning. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 15434–15443, 2023. 2

    2023

  11. [19]

    Using duck-net for polyp image segmentation

    Razvan-Gabriel Dumitru, Darius Peteleaza, and Catalin Craciun. Using duck-net for polyp image segmentation. Sci- entific reports, 13(1):9803, 2023. 2

    2023

  12. [20]

    Everingham, L

    M. Everingham, L. Van Gool, C. K. I. Williams, J. Winn, and A. Zisserman. The pascal visual object classes (voc) challenge. International Journal of Computer Vision, 88(2): 303–338, 2010. 2

    2010

  13. [21]

    Everingham, L

    M. Everingham, L. Van Gool, C. K. I. Williams, J. Winn, and A. Zisserman. The PASCAL Visual Object Classes Challenge 2012 (VOC2012) Results. http://www.pascal- network.org/challenges/VOC/voc2012/workshop/index.html,

    2012

  14. [22]

    How do training methods influence the utiliza- tion of vision models? arXiv preprint arXiv:2410.14470 ,

    Paul Gavrikov, Shashank Agnihotri, Margret Keuper, and Ja- nis Keuper. How do training methods influence the utiliza- tion of vision models? arXiv preprint arXiv:2410.14470 ,

    arXiv

  15. [23]

    Robust models are less over-confident

    Julia Grabinski, Paul Gavrikov, Janis Keuper, and Margret Keuper. Robust models are less over-confident. NeurIPS,

  16. [24]

    Frequencylowcut pooling-plug and play against catas- trophic overfitting

    Julia Grabinski, Steffen Jung, Janis Keuper, and Margret Ke- uper. Frequencylowcut pooling-plug and play against catas- trophic overfitting. In European Conference on Computer Vision, pages 36–57. Springer, 2022

    2022

  17. [25]

    Alias- ing and adversarial robust generalization of cnns

    Julia Grabinski, Janis Keuper, and Margret Keuper. Alias- ing and adversarial robust generalization of cnns. Machine Learning, pages 1–27, 2022. 2

    2022

  18. [26]

    Segpgd: An effective and efficient adversarial attack for evaluating and boosting segmentation robustness

    Jindong Gu, Hengshuang Zhao, V olker Tresp, and Philip HS Torr. Segpgd: An effective and efficient adversarial attack for evaluating and boosting segmentation robustness. In ECCV, pages 308–325. Springer, 2022. 10

    2022

  19. [27]

    Robustifying token attention for vision transformers

    Yong Guo, David Stutz, and Bernt Schiele. Robustifying token attention for vision transformers. In Proceedings of the IEEE/CVF International Conference on Computer Vi- sion, pages 17557–17568, 2023. 3 5

    2023

  20. [28]

    Semantic contours from inverse detectors

    Bharath Hariharan, Pablo Arbelaez, Lubomir Bourdev, Subhransu Maji, and Jitendra Malik. Semantic contours from inverse detectors. In International Conference on Computer Vision (ICCV), 2011. 10

    2011

  21. [29]

    Hypercolumns for object segmentation and fine-grained localization

    Bharath Hariharan, Pablo Arbel ´aez, Ross Girshick, and Ji- tendra Malik. Hypercolumns for object segmentation and fine-grained localization. In CVPR, pages 447–456, 2015. 10

    2015

  22. [30]

    Deep residual learning for image recognition

    Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. Deep residual learning for image recognition. In Proc. of the IEEE conference on computer vision and pattern recog- nition, pages 770–778, 2016. 10, 11

    2016

  23. [31]

    Benchmarking neu- ral network robustness to common corruptions and perturba- tions

    Dan Hendrycks and Thomas Dietterich. Benchmarking neu- ral network robustness to common corruptions and perturba- tions. In Proc. International Conference on Learning Repre- sentations (ICLR), 2019. 1, 2, 3, 4, 10, 13

    2019

  24. [32]

    Cubuk, Barret Zoph, Justin Gilmer, and Balaji Lakshminarayanan

    Dan Hendrycks, Norman Mu, Ekin D. Cubuk, Barret Zoph, Justin Gilmer, and Balaji Lakshminarayanan. AugMix: A simple data processing method to improve robustness and uncertainty. Proc. of the International Conference on Learn- ing Representations (ICLR), 2020. 2, 3

    2020

  25. [33]

    Towards improving robustness of compressed cnns

    J Hoffmann, S Agnihotri, Tonmoy Saikia, and Thomas Brox. Towards improving robustness of compressed cnns. InICML Workshop on Uncertainty and Robustness in Deep Learning (UDL), 2021. 2, 3

    2021

  26. [34]

    Mobilenets: Efficient convolu- tional neural networks for mobile vision applications

    Andrew G Howard, Menglong Zhu, Bo Chen, Dmitry Kalenichenko, Weijun Wang, Tobias Weyand, Marco An- dreetto, and Hartwig Adam. Mobilenets: Efficient convolu- tional neural networks for mobile vision applications. arXiv preprint arXiv:1704.04861, 2017. 10

    2017 arXiv

  27. [35]

    Benchmarking the robustness of semantic segmentation models

    Christoph Kamann and Carsten Rother. Benchmarking the robustness of semantic segmentation models. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 8828–8838, 2020. 2, 3, 9, 10

    2020

  28. [36]

    3d common corruptions and data augmentation

    O ˘guzhan Fatih Kar, Teresa Yeo, Andrei Atanov, and Amir Zamir. 3d common corruptions and data augmentation. In Proc. IEEE/CVF Conference on Computer Vision and Pat- tern Recognition (CVPR), pages 18963–18974, 2022. 2, 3, 10

    2022

  29. [37]

    Hierarchical markov random fields for mast cell segmentation in electron microscopic record- ings

    Margret Keuper, Thorsten Schmidt, Marta Rodriguez- Franco, Wolfgang Schamel, Thomas Brox, Hans Burkhardt, and Olaf Ronneberger. Hierarchical markov random fields for mast cell segmentation in electron microscopic record- ings. In 2011 IEEE International Symposium on Biomedical...

    2011

  30. [38]

    Intra-source style augmentation for improved domain gener- alization

    Yumeng Li, Dan Zhang, Margret Keuper, and Anna Khoreva. Intra-source style augmentation for improved domain gener- alization. In Proceedings of the IEEE/CVF Winter Confer- ence on Applications of Computer Vision , pages 509–519,

  31. [39]

    Adversarial supervision makes layout-to-image diffusion models thrive

    Yumeng Li, Margret Keuper, Dan Zhang, and Anna Khoreva. Adversarial supervision makes layout-to-image diffusion models thrive. In The Twelfth International Conference on Learning Representations, 2024. 2

    2024

  32. [40]

    Intra-& extra-source exemplar-based style synthesis for im- proved domain generalization

    Yumeng Li, Dan Zhang, Margret Keuper, and Anna Khoreva. Intra-& extra-source exemplar-based style synthesis for im- proved domain generalization. International Journal of Computer Vision, 132(2):446–465, 2024. 2

    2024

  33. [41]

    Microsoft coco: Common objects in context

    Tsung-Yi Lin, Michael Maire, Serge Belongie, James Hays, Pietro Perona, Deva Ramanan, Piotr Doll´ar, and C Lawrence Zitnick. Microsoft coco: Common objects in context. In Computer Vision–ECCV 2014: 13th European Conference, Zurich, Switzerland, September 6-12, 2014, Proceeding...

    2014

  34. [42]

    Swin transformer: Hierarchical vision transformer using shifted windows

    Ze Liu, Yutong Lin, Yue Cao, Han Hu, Yixuan Wei, Zheng Zhang, Stephen Lin, and Baining Guo. Swin transformer: Hierarchical vision transformer using shifted windows. In Proc. of the IEEE/CVF International Conference on Com- puter Vision (ICCV), 2021. 1, 11, 13

    2021

  35. [43]

    Towards class-wise robustness analysis

    Tejaswini Medi, Julia Grabinski, and Margret Keuper. Towards class-wise robustness analysis. arXiv preprint arXiv:2411.19853, 2024. 2

    2024 arXiv

  36. [44]

    Fair-tat: Improving model fairness using targeted adversarial train- ing

    Tejaswini Medi, Steffen Jung, and Margret Keuper. Fair-tat: Improving model fairness using targeted adversarial train- ing. In 2025 IEEE/CVF Winter Conference on Applications of Computer Vision (WACV), pages 7827–7836. IEEE, 2025. 2

    2025

  37. [45]

    Object scene flow for au- tonomous vehicles

    Moritz Menze and Andreas Geiger. Object scene flow for au- tonomous vehicles. In Proc. IEEE/CVF Conference on Com- puter Vision and Pattern Recognition (CVPR), pages 3061– 3070, 2015. 2

    2015

  38. [46]

    Ecker, Matthias Bethge, and Wieland Brendel

    Claudio Michaelis, Benjamin Mitzkus, Robert Geirhos, Evgenia Rusak, Oliver Bringmann, Alexander S. Ecker, Matthias Bethge, and Wieland Brendel. Benchmarking ro- bustness in object detection: Autonomous driving when win- ter is coming. arXiv preprint arXiv:1907.07484, 2019. 2, 3

    1907 arXiv

  39. [47]

    Clas- sification robustness to common optical aberrations

    Patrick M ¨uller, Alexander Braun, and Margret Keuper. Clas- sification robustness to common optical aberrations. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 3632–3643, 2023. 10

    2023

  40. [48]

    Aligning visual and seman- tic interpretability through visually grounded concept bottle- neck models

    Katharina Prasse, Patrick Knab, Sascha Marton, Christian Bartelt, and Margret Keuper. Aligning visual and seman- tic interpretability through visually grounded concept bottle- neck models. arXiv preprint arXiv:2412.11576, 2024. 2

    2024 arXiv

  41. [49]

    U-net: Convolutional networks for biomedical image segmentation

    Olaf Ronneberger, Philipp Fischer, and Thomas Brox. U-net: Convolutional networks for biomedical image segmentation. In MICCAI, pages 234–241. Springer, 2015. 2, 10, 11

    2015

  42. [50]

    ACDC: The adverse conditions dataset with correspondences for se- mantic driving scene understanding

    Christos Sakaridis, Dengxin Dai, and Luc Van Gool. ACDC: The adverse conditions dataset with correspondences for se- mantic driving scene understanding. In Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), 2021. 1, 2, 13

    2021

  43. [51]

    Detection defenses: An empty promise against adver- sarial patch attacks on optical flow

    Erik Scheurer, Jenny Schmalfuss, Alexander Lis, and Andr´es Bruhn. Detection defenses: An empty promise against adver- sarial patch attacks on optical flow. InProc. IEEE/CVF Win- ter Conference on Applications of Computer Vision (WACV),

  44. [52]

    At- tacking motion estimation with adversarial snow

    Jenny Schmalfuss, Lukas Mehl, and Andr ´es Bruhn. At- tacking motion estimation with adversarial snow. In Proc. ECCV Workshop on Adversarial Robustness in the Real World (AROW), 2022

    2022

  45. [53]

    A perturbation-constrained adversarial attack for evaluating the 6 robustness of optical flow

    Jenny Schmalfuss, Philipp Scholze, and Andr ´es Bruhn. A perturbation-constrained adversarial attack for evaluating the 6 robustness of optical flow. In Proc. European Conference on Computer Vision (ECCV), pages 183–200, 2022. 10

    2022

  46. [54]

    Implicit representations for constrained image segmentation

    Jan Philipp Schneider, Mishal Fatima, Jovita Lukasik, An- dreas Kolb, Margret Keuper, and Michael Moeller. Implicit representations for constrained image segmentation. In Pro- ceedings of the 41st International Conference on Machine Learning (ICML) , pages 43765–43790. MLResea...

  47. [55]

    Differentiable sensor layouts for end-to-end learn- ing of task-specific camera parameters

    Hendrik Sommerhoff, Shashank Agnihotri, Mohamed Saleh, Michael Moeller, Margret Keuper, and Andreas Kolb. Differentiable sensor layouts for end-to-end learn- ing of task-specific camera parameters. arXiv preprint arXiv:2304.14736, 2023. 2

    2023 arXiv

  48. [56]

    Task driven sensor layouts-joint optimiza- tion of pixel layout and network parameters

    Hendrik Sommerhoff, Shashank Agnihotri, Mohamed Saleh, Michael Moeller, Margret Keuper, Bhaskar Choubey, and Andreas Kolb. Task driven sensor layouts-joint optimiza- tion of pixel layout and network parameters. In 2024 IEEE International Conference on Computational Photography...

    2024

  49. [57]

    Internimage: Exploring large-scale vi- sion foundation models with deformable convolutions

    Wenhai Wang, Jifeng Dai, Zhe Chen, Zhenhang Huang, Zhiqi Li, Xizhou Zhu, Xiaowei Hu, Tong Lu, Lewei Lu, Hongsheng Li, et al. Internimage: Exploring large-scale vi- sion foundation models with deformable convolutions. In Proceedings of the IEEE/CVF conference on computer vi- si...

    2023

  50. [58]

    Unified perceptual parsing for scene understand- ing

    Tete Xiao, Yingcheng Liu, Bolei Zhou, Yuning Jiang, and Jian Sun. Unified perceptual parsing for scene understand- ing. In Proceedings of the European conference on computer vision (ECCV), pages 418–434, 2018. 11

    2018

  51. [59]

    Segformer: Simple and efficient design for semantic segmentation with transform- ers

    Enze Xie, Wenhai Wang, Zhiding Yu, Anima Anandkumar, Jose M Alvarez, and Ping Luo. Segformer: Simple and efficient design for semantic segmentation with transform- ers. Advances in neural information processing systems, 34: 12077–12090, 2021. 9, 11

    2021

  52. [60]

    Improving 2d feature representations by 3d-aware fine-tuning

    Yuanwen Yue, Anurag Das, Francis Engelmann, Siyu Tang, and Jan Eric Lenssen. Improving 2d feature representations by 3d-aware fine-tuning. In European Conference on Com- puter Vision, pages 57–74. Springer, 2024. 2

    2024

  53. [61]

    Hengshuang Zhao. semseg. https://github.com/ hszhao/semseg, 2019. 10

    2019

  54. [62]

    Pyramid scene parsing network

    Hengshuang Zhao, Jianping Shi, Xiaojuan Qi, Xiaogang Wang, and Jiaya Jia. Pyramid scene parsing network. In CVPR, pages 2881–2890, 2017. 3, 9, 10, 11

    2017

  55. [63]

    Semantic under- standing of scenes through the ade20k dataset

    Bolei Zhou, Hang Zhao, Xavier Puig, Tete Xiao, Sanja Fi- dler, Adela Barriuso, and Antonio Torralba. Semantic under- standing of scenes through the ade20k dataset. International Journal of Computer Vision, 127:302–321, 2019. 2, 9 7 Are Synthetic Corruptions A Reliable Proxy Fo...

    2019

  56. [64]

    proposes introducing a distribution shift in the input samples by perturbing images with a total of 15 synthetic corruptions that could occur in the real world. These corruptions include weather phenomena such as fog, and frost, digital corruptions such as jpeg compression, pi...

    2012

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.