Pith. sign in

REVIEW 1 major objections 4 minor 62 references

What is the Added Value of UDA in the VFM Era?

T0 review · 1 major / 4 minor · reviewed 2026-08-16 · deepseek-v4-flash

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

Pith's one-line read Under realistic data conditions, unsupervised domain adaptation adds little over simply fine-tuning a vision foundation model.

desk verdict UDA's added value over VFM fine-tuning shrinks to near zero when source data is scaled and diversified, but the paper's strongest conclusion rests on a single method and a -0.3 mIoU difference. read the letter →

arxiv 2504.18190 v1 pith:WFU7MJHH submitted 2025-04-25 cs.CV

Brunó B. Englert , Tommie Kerssies , Gijs Dubbelman This is my paper · ORCID
classification cs.CV
keywords unsuperviseddomainadaptationvisionfoundationmodelssemanticsegmentationsource-onlyfine-tuningsynthetic-to-realreal-to-realautonomousdrivingCityscapes
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 unsupervised domain adaptation (UDA) still earns its complexity now that vision foundation models (VFMs) already generalize strongly on their own. It compares UDA against plain source-only fine-tuning of the same VFM on semantic segmentation for autonomous driving, across synthetic-to-real and real-to-real scenarios with varying source diversity and small amounts of labeled target data. The central finding is that UDA's added value shrinks as source data becomes richer: from +8.0 mIoU to +1.8 mIoU with stronger synthetic sources, and to no gain (indeed -0.3 mIoU) with diverse real sources. UDA still consistently beats source-only fine-tuning in synthetic-data scenarios, and with 1/16 of Cityscapes labels it matches fully-supervised performance at 85.1 mIoU, yet the paper concludes that UDA is not a key enabler for autonomous driving because simple fine-tuning is practically as good.

What carries the argument

The evaluative machinery is a controlled comparison between VFM-UDA++ as a representative state-of-the-art UDA method and source-only fine-tuning of the identical architecture (DINOv2-L encoder with ViT-Adapter and BasicPyramid decoder), across systematically varied source and target compositions. The method combines EMA-teacher pseudo-labeling, a feature-distance loss that prevents forgetting of VFM pre-training, and masked image consistency, with an optional two-stage procedure that mixes in 1/16 Cityscapes labels. This setup lets the authors isolate what UDA adds over straightforward fine-tuning as source diversity, target scale, and label availability change.

What would settle it

An independent team implements a different state-of-the-art UDA method (for example, a DINOv2-based variant of DAFormer or MIC) and runs the same scenarios with diverse source data; if its added value over source-only fine-tuning remains above +5 mIoU on Cityscapes, the conclusion that UDA has little practical value under diverse data would not generalize.

Watch

Extended reading notes

Core claim

On the paper's own terms, the discovery is that the performance gap that justified UDA—improving generalization from a labeled source to an unlabeled target—largely disappears when the source data is representative of what an autonomous-driving company would actually have. Using VFM-UDA++ with a DINOv2 encoder, the authors find that replacing the single GTA5 source with GTA5, SYNTHIA, and UrbanSyn cuts UDA's advantage over source-only fine-tuning from +8.0 to +1.8 mIoU on Cityscapes. In the real-to-real setting, moving from BDD alone to BDD, Mapillary Vistas, and ACDC turns a +2.6 mIoU UDA gain into -0.3 mIoU. The one consistently positive role for UDA appears when the source composition is less favorable: swapping UrbanSyn for SynScapes drops source-only fine-tuning by 3.8 mIoU while UDA stays robust, producing a 6.3 mIoU advantage. The paper therefore concludes that UDA's practical value in autonomous driving is as a targeted fallback, not a standard training paradigm.

Load-bearing premise

The study assumes that the single unsupervised-domain-adaptation method it tests is representative of the whole class, even though that method was developed by the same authors and is the only one used in the experiments.

Editorial extensions

If this is right

  • In synth-to-real pipelines, UDA retains a consistent but modest edge over source-only fine-tuning, and that edge grows to +6.3 mIoU when the source composition is suboptimal.
  • In real-to-real pipelines with diverse labeled source data, UDA no longer improves target accuracy over source-only fine-tuning (-0.3 mIoU), so adaptation buys robustness on WildDash2 (+0.8 mIoU) but not Cityscapes accuracy.
  • Scaling unlabeled target data, even with same-distribution Cityscapes extra data, does not improve target-domain generalization, aligning with UDA-Bench.
  • With only 1/16 of Cityscapes labels, two-stage UDA reaches 85.1 mIoU, equal to fully-supervised training on all labels, while source-only fine-tuning with the same labels reaches 83.0 mIoU.
  • The paper's conclusion is that UDA is not a key enabler for autonomous driving; source-only fine-tuning of VFMs achieves practically similar results in realistic settings.

Reading between the lines

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

  • A testable extension the paper leaves implicit: compare two-stage UDA against plain semi-supervised fine-tuning with the same 1/16 labels and no synthetic source, to separate UDA's contribution from label-efficient VFM fine-tuning.
  • If the pattern holds across other VFMs, the practical bottleneck shifts from adaptation algorithms to the acquisition of diverse labeled source data and the curation of source composition.
  • Because the single UDA implementation is from the same group, the 'no added value' results are most safely read as a statement about this method class; replication with independent implementations would raise confidence.
  • The WildDash2 robustness results suggest UDA acts as a regularizer against source-distribution shifts, which could motivate using UDA-like target-domain consistency even when labels are available.
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

1 major / 4 minor

Summary. The paper asks whether Unsupervised Domain Adaptation (UDA) still adds value in the era of Vision Foundation Models (VFMs), focusing on semantic segmentation for autonomous driving. Using VFM-UDA++ as the representative UDA method, the authors compare UDA against source-only fine-tuning across synthetic-to-real and real-to-real scenarios while scaling and diversifying source and target datasets, and also study the effect of adding a small amount (1/16) of labeled target data. The main findings are: (i) with stronger synthetic source data, UDA's improvement over source-only fine-tuning drops from +8.0 to +1.8 mIoU on Cityscapes (Table 2); (ii) scaling unlabeled target data has little or no effect (Table 3); (iii) UDA is less sensitive to changes in synthetic source composition than source-only fine-tuning (Table 4); (iv) in real-to-real settings with diverse labeled source data, UDA shows no added value, with a small negative difference of -0.3 mIoU (Table 6); and (v) with 1/16 of Cityscapes labels, UDA matches fully-supervised performance (Tables 5 and 7). The paper concludes that UDA is not a key enabler for autonomous driving, except as a fallback when domain gaps are substantial and labeled target data is unavailable.

Significance. If its conclusions hold, the paper provides a valuable and timely empirical reassessment of UDA in the VFM era. It is one of the few studies that systematically compares UDA with a strong source-only baseline across a variety of source-data compositions and realistic data scales, and it evaluates forgetting on WildDash2 in addition to Cityscapes, which is a useful methodological addition. The observation that scaling unlabeled target data provides little benefit corroborates results from UDA-Bench. However, the paper's broad negative conclusion about UDA's practical value rests on a single UDA implementation and on small, single-run performance differences, so the strength of the conclusion currently exceeds what the evidence can support. With additional independent UDA methods and uncertainty quantification, this could become a reference benchmark for the community.

major comments (1)
  1. [Sec. 4.4, Table 6; Sec. 4.3, Tables 2-4; Sec. 5] This comment is about the paper's treatment of the few-label results in Tables 5 and 7. The claim that UDA 'matches fully-supervised performance' with 1/16 labels is supported by a single run and the differences are small (85.1 vs 85.1 in Table 5 and 84.7 vs 84.7 in Table 7). Moreover, the fully-supervised baseline uses all Cityscapes labels but not the pseudo-labeling and consistency machinery, so the comparison has the same asymmetry issue noted above. The paper should report variability across seeds and clarify whether the matching is within noise. The positive claim that UDA uses small labeled target data better than source-only fine-tuning is plausible, but the current evidence does not justify the strength of the language.
minor comments (4)
  1. [Sec. 1] In Sec. 4.3, the sentence 'where the reference, where the reference setup (GTA5 → CS) is marked in gray' contains a duplicated phrase and should be rewritten.
  2. [Sec. 4.2] The paper does not mention whether code or configuration files will be released; given the empirical nature of the study and the emphasis on reproducibility, a statement on code/data release would be beneficial.
  3. [Sec. 3.4] The two-stage procedure for 'UDA with few target labels' is described as using the same UDA pipeline in a semi-supervised fashion, but the semi-supervised setting in Sec. 3.1 is defined as having no domain gap between labeled and unlabeled data; this distinction could be made more explicit.
  4. [General] Some datasets are abbreviated without explicit introduction at first use (e.g., S, US, SS in Tables 2-4); a short legend or note listing GTA5, SYNTHIA, UrbanSyn, SynScapes would improve readability.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the study is an empirical benchmark; UDA added value is measured, not derived from the method's definition.

full rationale

The central comparison is an ablation-style benchmark: UDA (VFM-UDA++) and source-only fine-tuning share the same architecture, data, and training budget, so the reported differences in mIoU are measured quantities rather than consequences of how the method is defined. No parameter is fitted to the outcome being explained, and no 'prediction' is generated from a fitted input; the added-value numbers in Tables 2-7 are direct experimental outputs of this paper. The use of the authors' own VFM-UDA++ as the representative UDA method is the only self-reference, and it is transparently justified by Table 1; relying on a single method affects external validity/generalizability, not logical circularity, because the conclusion would stand or fall with the measured numbers even if the method came from another group. The Discussion explicitly qualifies the conclusion by the diversity assumption, further showing the claim is conditional on empirical data rather than enforced by the setup. The paper is therefore self-contained against external benchmarks and contains no circular reduction by construction.

Assumptions & free parameters 0 free parameters · 5 assumptions · 0 invented entities

The central claim is not a derivation, so there are no mathematical free parameters fitted to data. The key choices that affect the conclusions are the selection of a single self-authored UDA method, the datasets used as target and out-of-target checks, the 1/16 label split, and the assumption that segmentation is representative.

assumptions (5)
  • domain assumption VFM-UDA++ is a representative state-of-the-art UDA method for evaluating UDA's added value.
    Only this method is used in all experiments; conclusions about 'UDA' generalize only if this method is representative. Invoked in Sec. 2 where the authors select a strong representative method and use their own VFM-UDA++.
  • domain assumption Cityscapes is a suitable focused target domain and WildDash2 is a valid out-of-target generalization check.
    Target and forgetting metrics are measured on these two datasets only; the conclusions about maintaining generalization depend on this choice. Described in Sec. 4.2.
  • domain assumption The 1/16 Cityscapes split represents a realistic small amount of labeled target data.
    Used for all few-label experiments; the claim that UDA matches fully-supervised with few labels is tied to this split. Described in Sec. 4.2.
  • domain assumption Semantic segmentation is representative of perception tasks relevant to autonomous driving.
    The study only evaluates semantic segmentation; the practical conclusions for AD assume transferability to other perception tasks. Stated at start of Sec. 4.
  • domain assumption In mainstream autonomous driving, labeled source data is sufficiently diverse to cover target domains.
    The conclusion that UDA is not a key enabler assumes diversity of source data; the authors themselves flag this assumption in Sec. 5 as difficult to guarantee.

how reviews work

0 comments
Cite this review

Pith. "Pith review of What is the Added Value of UDA in the VFM Era?." pith.science (2026). https://pith.science/paper/WFU7MJHH

@misc{pith2026250418190,
  author       = {Pith},
  title        = {Pith review of: What is the Added Value of UDA in the VFM Era?},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/WFU7MJHH}},
  note         = {Machine review of arXiv:2504.18190}
}
read the original abstract

Unsupervised Domain Adaptation (UDA) can improve a perception model's generalization to an unlabeled target domain starting from a labeled source domain. UDA using Vision Foundation Models (VFMs) with synthetic source data can achieve generalization performance comparable to fully-supervised learning with real target data. However, because VFMs have strong generalization from their pre-training, more straightforward, source-only fine-tuning can also perform well on the target. As data scenarios used in academic research are not necessarily representative for real-world applications, it is currently unclear (a) how UDA behaves with more representative and diverse data and (b) if source-only fine-tuning of VFMs can perform equally well in these scenarios. Our research aims to close these gaps and, similar to previous studies, we focus on semantic segmentation as a representative perception task. We assess UDA for synth-to-real and real-to-real use cases with different source and target data combinations. We also investigate the effect of using a small amount of labeled target data in UDA. We clarify that while these scenarios are more realistic, they are not necessarily more challenging. Our results show that, when using stronger synthetic source data, UDA's improvement over source-only fine-tuning of VFMs reduces from +8 mIoU to +2 mIoU, and when using more diverse real source data, UDA has no added value. However, UDA generalization is always higher in all synthetic data scenarios than source-only fine-tuning and, when including only 1/16 of Cityscapes labels, synthetic UDA obtains the same state-of-the-art segmentation quality of 85 mIoU as a fully-supervised model using all labels. Considering the mixed results, we discuss how UDA can best support robust autonomous driving at scale.

Figures

Figures reproduced from arXiv: 2504.18190 by the authors.

Figure 1
Figure 1. UDA methods vs. source-only baselines and fully￾supervised oracles. While VFM-based UDA achieves general￾ization close to fully-supervised learning, the added value of UDA over simple source-only fine-tuning requires further investigation. tem, vision models provide an autonomous vehicle with the crucial ability to perceive and understand its surroundings. Developing vision models that are robust to diverse data dis… view at source ↗
Figure 2
Figure 2. Dataset Overview. On the top left, we show source-only performance (mIoU) evaluated on Cityscapes [9] for models trained on different source datasets, including their domain gaps relative to the oracle (trained on Cityscapes), and dataset sizes (in thousands of samples). On the top right, a t-SNE visualization of DINOv2-L [32] [CLS] token embeddings shows a clear separation between synthetic and real datasets, while… view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

62 extracted references · 53 canonical work pages

  1. [1]

    Self-supervised Augmen- tation Consistency for Adapting Semantic Segmentation

    Nikita Araslanov and Stefan Roth. Self-supervised Augmen- tation Consistency for Adapting Semantic Segmentation. In CVPR, 2021. 3

    work page 2021

  2. [2]

    PASTA: Proportional Ampli- tude Spectrum Training Augmentation for Syn-to-Real Do- main Generalization

    Prithvijit Chattopadhyay, Kartik Sarangmath, Vivek Vi- jaykumar, and Judy Hoffman. PASTA: Proportional Ampli- tude Spectrum Training Augmentation for Syn-to-Real Do- main Generalization. In CVPR, 2023. 3

    work page 2023

  3. [3]

    End-to-end Autonomous Driving: Challenges and Frontiers

    Li Chen, Penghao Wu, Kashyap Chitta, Bernhard Jaeger, An- dreas Geiger, and Hongyang Li. End-to-end Autonomous Driving: Challenges and Frontiers. TPAMI, 2024. 1

    work page 2024

  4. [4]

    PiPa: Pixel- and Patch-wise Self-supervised Learning for Domain Adaptative Semantic Segmentation

    Mu Chen, Zhedong Zheng, Yi Yang, and Tat-Seng Chua. PiPa: Pixel- and Patch-wise Self-supervised Learning for Domain Adaptative Semantic Segmentation. ACM, 2023. 2

    work page 2023

  5. [5]

    Transferring to Real-World Layouts: A Depth-aware Framework for Scene Adaptation

    Mu Chen, Zhedong Zheng, and Yi Yang. Transferring to Real-World Layouts: A Depth-aware Framework for Scene Adaptation. In ACM Multimedia, 2024. 2, 3

    work page 2024

  6. [6]

    Vision Transformer Adapter for Dense Predictions

    Zhe Chen, Yuchen Duan, Wenhai Wang, Junjun He, Tong Lu, Jifeng Dai, and Yu Qiao. Vision Transformer Adapter for Dense Predictions. In ICLR, 2023. 2, 5

    work page 2023

  7. [7]

    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,

    work page

  8. [8]

    Robustnet: Improving domain generalization in urban-scene segmentation via in- stance selective whitening

    Sungha Choi, Sanghun Jung, Huiwon Yun, Joanne T Kim, Seungryong Kim, and Jaegul Choo. Robustnet: Improving domain generalization in urban-scene segmentation via in- stance selective whitening. In CVPR, 2021. 3

    work page 2021

Show all 62 references
  1. [9]

    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 CVPR,

  2. [10]

    Dark model adaptation: Semantic image segmentation from daytime to nighttime

    Dengxin Dai and Luc Van Gool. Dark model adaptation: Semantic image segmentation from daytime to nighttime. In ITSC, 2018. 6

    2018

  3. [11]

    ImageNet: A large-scale hierarchical image database

    Jia Deng, Wei Dong, Richard Socher, Li-Jia Li, Kai Li, and Li Fei-Fei. ImageNet: A large-scale hierarchical image database. In CVPR, 2009. 3

    2009

  4. [12]

    and Dubbelman, Gijs

    Englert, Brun ´o B. and Dubbelman, Gijs. VFM-UDA++: Im- proving Network Architectures and Data Strategies for Un- supervised Domain Adaptive Semantic Segmentation. arXiv preprint arXiv:2503.10685, 2025. 2, 3, 4, 5, 6

    2025 arXiv

  5. [13]

    Exploring the Benefits of Vision Foundation Models for Unsupervised Domain Adap- tation

    Englert, Brun ´o B., Piva, Fabrizio J., Kerssies, Tommie, de Geus, Daan, and Dubbelman, Gijs. Exploring the Benefits of Vision Foundation Models for Unsupervised Domain Adap- tation. In CVPRW, 2024. 2, 3, 6

    2024

  6. [14]

    EV A-02: A Visual Representation for Neon Genesis

    Yuxin Fang, Quan Sun, Xinggang Wang, Tiejun Huang, Xin- long Wang, and Yue Cao. EV A-02: A Visual Representation for Neon Genesis. arXiv preprint arXiv:2303.11331, 2023. 2

    2023 arXiv

  7. [15]

    Finlayson

    Geoffrey French, Timo Aila, Samuli Laine, Michal Mack- iewicz, and Graham D. Finlayson. Consistency regulariza- tion and CutMix for semi-supervised semantic segmentation. CoRR, abs/1906.01916, 2019. 4

    1906 arXiv

  8. [16]

    G ´omez, Manuel Silva, Antonio Seoane, Agn `es Borr´as, Mario Noriega, Germ ´an Ros, Jose A

    Jose L. G ´omez, Manuel Silva, Antonio Seoane, Agn `es Borr´as, Mario Noriega, Germ ´an Ros, Jose A. Iglesias- Guitian, and Antonio M. L ´opez. All for One, and One for All: UrbanSyn Dataset, the third Musketeer of Synthetic Driving Scenes, 2023. 6, 7

    2023

  9. [17]

    Girshick

    Kaiming He, Xinlei Chen, Saining Xie, Yanghao Li, Piotr Doll´ar, and Ross B. Girshick. Masked Autoencoders Are Scalable Vision Learners. In CVPR, 2022. 2

    2022

  10. [18]

    FCNs in the Wild: Pixel-level Adversarial and Constraint- based Adaptation

    Judy Hoffman, Dequan Wang, Fisher Yu, and Trevor Darrell. FCNs in the Wild: Pixel-level Adversarial and Constraint- based Adaptation. CoRR, abs/1612.02649, 2016. 1

    2016 arXiv

  11. [19]

    Cycada: Cycle-consistent adversarial domain adaptation

    Judy Hoffman, Eric Tzeng, Taesung Park, Jun-Yan Zhu, Phillip Isola, Kate Saenko, Alexei Efros, and Trevor Darrell. Cycada: Cycle-consistent adversarial domain adaptation. In ICML, pages 1989–1998,. 2018. 3

    1989

  12. [20]

    Beyond Pixels: Semi-Supervised Semantic Segmenta- tion with a Multi-scale Patch-based Multi-Label Classifier

    Prantik Howlader, Srijan Das, Hieu Le, and Dimitris Sama- ras. Beyond Pixels: Semi-Supervised Semantic Segmenta- tion with a Multi-scale Patch-based Multi-Label Classifier. arXiv preprint arXiv:2407.04036, 2024. 2

    2024 arXiv

  13. [21]

    DAFormer: Improving network architectures and training strategies for domain-adaptive semantic segmentation

    Lukas Hoyer, Dengxin Dai, and Luc Gool. DAFormer: Improving network architectures and training strategies for domain-adaptive semantic segmentation. In CVPR, 2022. 3, 4

    2022

  14. [22]

    HRDA: Context- aware high-resolution domain-adaptive semantic segmenta- tion

    Lukas Hoyer, Dengxin Dai, and Luc Gool. HRDA: Context- aware high-resolution domain-adaptive semantic segmenta- tion. In ECCV, 2022. 2, 3

    2022

  15. [23]

    MIC: Masked Image Consistency for Context- Enhanced Domain Adaptation

    Lukas Hoyer, Dengxin Dai, Haoran Wang, and Luc Van Gool. MIC: Masked Image Consistency for Context- Enhanced Domain Adaptation. In CVPR, 2023. 2, 3, 4

    2023

  16. [24]

    SemiVL: Semi- Supervised Semantic Segmentation with Vision-Language Guidance

    Lukas Hoyer, David Joseph Tan, Muhammad Ferjad Naeem, Luc Van Gool, and Federico Tombari. SemiVL: Semi- Supervised Semantic Segmentation with Vision-Language Guidance. In ECCV, 2024. 2, 6

    2024

  17. [25]

    UDA-Bench: Revisiting Common Assumptions 9 in Unsupervised Domain Adaptation Using a Standardized Framework

    Tarun Kalluri, Sreyas Ravichandran, and Manmohan Chan- draker. UDA-Bench: Revisiting Common Assumptions 9 in Unsupervised Domain Adaptation Using a Standardized Framework. In ECCV, 2024. 7

    2024

  18. [26]

    How to Benchmark Vision Foundation Models for Semantic Seg- mentation? In CVPRW, 2024

    Tommie Kerssies, Daan de Geus, and Gijs Dubbelman. How to Benchmark Vision Foundation Models for Semantic Seg- mentation? In CVPRW, 2024. 3

    2024

  19. [27]

    First Place Solution to the ECCV 2024 BRA VO Challenge: Eval- uating Robustness of Vision Foundation Models for Seman- tic Segmentation

    Tommie Kerssies, Daan de Geus, and Gijs Dubbelman. First Place Solution to the ECCV 2024 BRA VO Challenge: Eval- uating Robustness of Vision Foundation Models for Seman- tic Segmentation. arXiv preprint arXiv:2409.17208, 2024

    2024 arXiv

  20. [28]

    Your ViT is Secretly an Image Segmentation Model

    Tommie Kerssies, Niccol `o Cavagnero, Alexander Hermans, Narges Norouzi, Giuseppe Averta, Bastian Leibe, Gijs Dubbelman, and Daan de Geus. Your ViT is Secretly an Image Segmentation Model. In CVPR, 2025. 3

    2025

  21. [29]

    Temporal Ensembling for Semi-Supervised Learning

    Samuli Laine and Timo Aila. Temporal Ensembling for Semi-Supervised Learning. In ICLR, 2017. 3

    2017

  22. [30]

    Decoupled weight decay regularization

    Ilya Loshchilov and Frank Hutter. Decoupled weight decay regularization. In ICLR, 2019. 5

    2019

  23. [31]

    The Mapillary Vistas Dataset for Seman- tic Understanding of Street Scenes

    Gerhard Neuhold, Tobias Ollmann, Samuel Rota Bul `o, and Peter Kontschieder. The Mapillary Vistas Dataset for Seman- tic Understanding of Street Scenes. In ICCV, 2017. 6

    2017

  24. [32]

    Rabbat, Vasu Sharma, Gabriel Synnaeve, Hu Xu, Herv´e J´egou, Julien Mairal, Patrick Labatut, Armand Joulin, and Piotr Bojanowski

    Maxime Oquab, Timoth ´ee Darcet, Th ´eo Moutakanni, Huy V o, Marc Szafraniec, Vasil Khalidov, Pierre Fernandez, Daniel Haziza, Francisco Massa, Alaaeldin El-Nouby, Mah- moud Assran, Nicolas Ballas, Wojciech Galuba, Rus- sell Howes, Po-Yao Huang, Shang-Wen Li, Ishan Misra, Mich...

    2024

  25. [33]

    Two at once: Enhancing learning and generalization capacities via ibn-net

    Xingang Pan, Ping Luo, Jianping Shi, and Xiaoou Tang. Two at once: Enhancing learning and generalization capacities via ibn-net. In ECCV, pages 464–479, 2018. 3

    2018

  26. [34]

    Global and local texture randomization for synthetic-to-real semantic segmentation

    Duo Peng, Yinjie Lei, Lingqiao Liu, Pingping Zhang, and Jun Liu. Global and local texture randomization for synthetic-to-real semantic segmentation. IEEE Transactions on Image Processing, 30:6594–6608, 2021. 3

    2021

  27. [35]

    Diffusion- based Image Translation with Label Guidance for Domain Adaptive Semantic Segmentation

    Duo Peng, Ping Hu, Qiuhong Ke, and Jun Liu. Diffusion- based Image Translation with Label Guidance for Domain Adaptive Semantic Segmentation . In ICCV, pages 808–820, Los Alamitos, CA, USA, 2023. IEEE Computer Society. 2

    2023

  28. [36]

    Piva, Daan de Geus, and Gijs Dubbelman

    Fabrizio J. Piva, Daan de Geus, and Gijs Dubbelman. Em- pirical Generalization Study: Unsupervised Domain Adapta- tion vs. Domain Generalization Methods for Semantic Seg- mentation in the Wild. In WACV, 2023. 6

    2023

  29. [37]

    Learning Transferable Visual Models From Natural Language Supervision

    Alec Radford, Jong Wook Kim, Chris Hallacy, Aditya Ramesh, Gabriel Goh, Sandhini Agarwal, Girish Sastry, Amanda Askell, Pamela Mishkin, Jack Clark, Gretchen Krueger, and Ilya Sutskever. Learning Transferable Visual Models From Natural Language Supervision. In ICML,

  30. [38]

    Richter, Vibhav Vineet, Stefan Roth, and Vladlen Koltun

    Stephan R. Richter, Vibhav Vineet, Stefan Roth, and Vladlen Koltun. Playing for Data: Ground Truth from Computer Games. In ECCV, 2016. 3, 6

    2016

  31. [39]

    Meletis, and G

    Rob Romijnders, P. Meletis, and G. Dubbelman. A Domain Agnostic Normalization Layer for Unsupervised Adversarial Domain Adaptation. In WACV, pages 1866–1875, 2019. 3

    2019

  32. [40]

    German Ros, Laura Sellart, Joanna Materzynska, David Vazquez, and Antonio M. Lopez. The SYNTHIA Dataset: A Large Collection of Synthetic Images for Semantic Seg- mentation of Urban Scenes. In CVPR, 2016. 3, 6

    2016

  33. [41]

    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 ICCV, pages 10745– 10755. IEEE, 2021. 6

    2021

  34. [42]

    DiGA: Distil to generalize and then adapt for domain adaptive semantic segmentation

    Fengyi Shen, Akhil Gurram, Ziyuan Liu, He Wang, and Alois Knoll. DiGA: Distil to generalize and then adapt for domain adaptive semantic segmentation. In CVPR, pages 15866–15877, 2023. 2

    2023

  35. [43]

    Fixmatch: Simplifying semi- supervised learning with consistency and confidence

    Kihyuk Sohn, David Berthelot, Chun-Liang Li, Zizhao Zhang, Nicholas Carlini, Ekin D Cubuk, Alex Kurakin, Han Zhang, and Colin Raffel. Fixmatch: Simplifying semi- supervised learning with consistency and confidence. In NeurIPS, 2020. 3

    2020

  36. [44]

    Zhiqiang Tang, Yunhe Gao, Yi Zhu, Zhi Zhang, Mu Li, and Dimitris N. Metaxas. CrossNorm and SelfNorm for Gener- alization Under Distribution Shifts. In ICCV, pages 52–61,

  37. [45]

    Mean teachers are better role models: Weight-averaged consistency targets improve semi-supervised deep learning results

    Antti Tarvainen and Harri Valpola. Mean teachers are better role models: Weight-averaged consistency targets improve semi-supervised deep learning results. In NeurIPS, pages 1195–1204,. 2017. 4

    2017

  38. [46]

    DACS: Domain Adaptation via Crossdo- main Mixed Sampling

    Wilhelm Tranheden, Viktor Olsson, Juliano Pinto, and Lennart Svensson. DACS: Domain Adaptation via Crossdo- main Mixed Sampling. In WACV, pages 1379–1389,. 2021. 3

    2021

  39. [47]

    CLUDA: Contrastive Learn- ing in Unsupervised Domain Adaptation for Semantic Seg- mentation

    Midhun Vayyat, Jaswin Kasi, Anuraag Bhattacharya, Shuaib Ahmed, and Rahul Tallamraju. CLUDA: Contrastive Learn- ing in Unsupervised Domain Adaptation for Semantic Seg- mentation. arXiv preprint arXiv:2208.14227, 2022. 3

    2022 arXiv

  40. [48]

    ADVENT: Adversarial Entropy Minimization for Domain Adaptation in Semantic Segmen- tation

    Tuan-Hung Vu, Himalaya Jain, Maxime Bucher, Mathieu Cord, and Patrick P ´erez. ADVENT: Adversarial Entropy Minimization for Domain Adaptation in Semantic Segmen- tation. In CVPR, 2019. 3

    2019

  41. [49]

    The BRA VO Semantic Segmentation Challenge Results in UNCV2024

    Tuan-Hung Vu, Eduardo Valle, Andrei Bursuc, Tommie Kerssies, Daan de Geus, Gijs Dubbelman, Long Qian, Bingke Zhu, Yingying Chen, Ming Tang, Jinqiao Wang, Tom´aˇs V oj´ıˇr, Jan ˇSochman, Jiˇr´ı Matas, Michael Smith, Frank Ferrie, Shamik Basu, Christos Sakaridis, and Luc Van Goo...

    2024

  42. [50]

    Harnessing diffusion models for visual perception with meta prompts

    Qiang Wan, Ming Nie, Zilong Huang, Bingyi Kang, Ji- ashi Feng, and Li Zhang. Harnessing diffusion models for visual perception with meta prompts. arXiv preprint arXiv:2312.14733, 2023. 2

    2023 arXiv

  43. [51]

    Allspark: Reborn labeled features from unlabeled in trans- former for semi-supervised semantic segmentation

    Haonan Wang, Qixiang Zhang, Yi Li, and Xiaomeng Li. Allspark: Reborn labeled features from unlabeled in trans- former for semi-supervised semantic segmentation. In CVPR, 2024. 2

    2024

  44. [52]

    CDAC: Cross-domain Attention Consistency in Transformer for Domain Adaptive Semantic Segmentation

    Kaihong Wang, Donghyun Kim, Rog ´erio Feris, and Margrit Betke. CDAC: Cross-domain Attention Consistency in Transformer for Domain Adaptive Semantic Segmentation. pages 11485–11495, 2023. 2

    2023

  45. [53]

    10 Stronger, Fewer, & Superior: Harnessing Vision Foundation Models for Domain Generalized Semantic Segmentation

    Zhixiang Wei, Lin Chen, Yi Jin, Xiaoxiao Ma, Tianle Liu, Pengyang Ling, Ben Wang, Huaian Chen, and Jinjin Zheng. 10 Stronger, Fewer, & Superior: Harnessing Vision Foundation Models for Domain Generalized Semantic Segmentation. In CVPR, 2024. 2, 3

    2024

  46. [54]

    Synscapes: A Photo- realistic Synthetic Dataset for Street Scene Parsing

    Magnus Wrenninge and Jonas Unger. Synscapes: A Photo- realistic Synthetic Dataset for Street Scene Parsing. CoRR, abs/1810.08705, 2018. 6, 7

    2018 arXiv

  47. [55]

    SePiCo: Semantic-Guided Pixel Contrast for Domain Adaptive Semantic Segmentation

    Binhui Xie, Shuang Li, Mingjia Li, Chi Harold Liu, Gao Huang, and Guoren Wang. SePiCo: Semantic-Guided Pixel Contrast for Domain Adaptive Semantic Segmentation. IEEE TPAMI, 45(07):9004–9021, 2023. 2, 3

    2023

  48. [56]

    ST++: Make Self-training Work Better for Semi-supervised Semantic Segmentation

    Lihe Yang, Wei Zhuo, Lei Qi, Yinghuan Shi, and Yang Gao. ST++: Make Self-training Work Better for Semi-supervised Semantic Segmentation. In CVPR, 2022. 3

    2022

  49. [57]

    Revisiting Weak-to-Strong Consistency in Semi-Supervised Semantic Segmentation

    Lihe Yang, Lei Qi, Litong Feng, Wayne Zhang, and Yinghuan Shi. Revisiting Weak-to-Strong Consistency in Semi-Supervised Semantic Segmentation. In CVPR, 2023. 3

    2023

  50. [58]

    Unimatch v2: Pushing the limit of semi-supervised semantic segmentation

    Lihe Yang, Zhen Zhao, and Hengshuang Zhao. Unimatch v2: Pushing the limit of semi-supervised semantic segmentation. TPAMI, 2025. 2, 3

    2025

  51. [59]

    BDD100K: A Diverse Driving Dataset for Heteroge- neous Multitask Learning

    Fisher Yu, Haofeng Chen, Xin Wang, Wenqi Xian, Yingying Chen, Fangchen Liu, Vashisht Madhavan, and Trevor Dar- rell. BDD100K: A Diverse Driving Dataset for Heteroge- neous Multitask Learning. In CVPR, 2020. 6

    2020

  52. [60]

    A Simple Baseline for Semi-supervised Semantic Segmentation with Strong Data Augmentation

    Jianlong Yuan, Yifan Liu, Chunhua Shen, Zhibin Wang, and Hao Li. A Simple Baseline for Semi-supervised Semantic Segmentation with Strong Data Augmentation. In ICCV,

  53. [61]

    WildDash - Creating Hazard-Aware Benchmarks

    Oliver Zendel, Katrin Honauer, Markus Murschitz, Daniel Steininger, and Gustavo Fern´andez Dom´ınguez. WildDash - Creating Hazard-Aware Benchmarks. In ECCV, 2018. 6

    2018

  54. [62]

    Adversarial style augmentation for domain general- ized urban-scene segmentation

    Zhun Zhong, Yuyang Zhao, Gim Hee Lee, and Nicu Sebe. Adversarial style augmentation for domain general- ized urban-scene segmentation. NeurIPS, 35:338–350, 2022. 3 11

    2022

Pith tools

Reviewed August 16, 2026 · model on record in the stance chip above.