Pith. sign in

REVIEW 4 major objections 5 minor 38 references

IDAL: Improved Domain Adaptive Learning for Natural Images Dataset

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 unsupervised domain adaptation for natural images improves when alignment is conditioned on pseudo-labels and the backbone sees multiple scales, beating CNN-based state of the art on two benchmarks and matching it on…

desk verdict A plausible incremental UDA method whose headline SOTA claim is undercut by the paper's own tables, especially on Office-31. read the letter →

arxiv 2506.17931 v1 pith:SVN2VUKW submitted 2025-06-22 cs.CV cs.AIcs.LG

Ravi Kant Gupta , Shounak Das , Amit Sethi This is my paper · ORCID
classification cs.CVcs.AIcs.LG
keywords unsuperviseddomainadaptationpseudo-labelmaximummeandiscrepancyfeaturepyramidnetworkconditionaladversarialnaturalimageclassificationshiftOffice-HomebenchmarkVisDA-2017
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

IDAL is a recipe for unsupervised domain adaptation (UDA) on natural images: take a ResNet-50 backbone, add a feature pyramid network (FPN), and train under a conditional adversarial objective with four losses, including a new pseudo-label maximum mean discrepancy (PLMMD) loss. The paper's central claim is that weighting MMD's kernel expectations by class pseudo-labels makes domain alignment class-aware, so target classes cluster while domain differences disappear. On the paper's experiments, this beats previous CNN-based UDA methods on Office-Home and VisDA-2017 and is comparable on Office-31 and DomainNet. The authors also report faster convergence, with useful target-domain structure appearing within a few epochs. A sympathetic reader would care because the gains are obtained with a moderate CNN and standard building blocks rather than a large transformer.

What carries the argument

The central machinery is the pseudo-label maximum mean discrepancy loss, a class-weighted version of the MMD two-sample test: for each kernel expectation in the MMD statistic, the contribution of each pair of instances is scaled by a weight derived from their pseudo-label vectors, so alignment is driven by matching the same class across domains rather than by matching marginal distributions. The supporting machinery is the ResNet-50 plus FPN feature extractor, which produces multi-scale features so object scale and style variations are separated, and the conditional adversarial domain adaptation (CDAN) objective, whose discriminator is conditioned on classifier predictions. The full training objective in Eq. (5) combines the source classification loss, the adversarial discriminator loss, and four penalty terms with hyper-parameters $\beta$, $\gamma$, $\delta$, and $\eta$ balancing information maximization, minimum class confusion, MMD, and PLMMD.

What would settle it

Train IDAL on Office-Home exactly as reported but with PLMMD weights computed from shuffled or randomly assigned pseudo-labels. If average accuracy stays near the reported 73.5% while removing PLMMD entirely drops it only to 73.2%, then the class-conditioning weights are not the active ingredient; if accuracy falls instead, the pseudo-label mechanism is doing the claimed work.

Watch

Extended reading notes

Core claim

On the paper's own terms, the discovery is that adversarial domain alignment benefits from being explicitly conditioned on class information at two levels: the FPN separates image scales so content and style information can be handled at different resolutions, and the PLMMD loss modifies the standard MMD test so that kernel expectations are weighted by pseudo-labels generated from the target classifier after a few iterations. In Eq. (9), $L_{PLMMD}$ multiplies the source, target, and cross-domain kernel expectations by weights $w_{XX}$, $w_{YY}$, and $w_{XY}$, respectively. These weights are described as normalized dot products of class-label vectors divided by the number of common classes, intended to up-weight within-class and cross-class pairs while down-weighting pairs from different classes. The authors claim that this class-conditional alignment, combined with information maximization, MMD, and minimum class confusion losses, yields better target accuracy and robustness than previous CNN-based UDA methods and speeds up convergence.

Load-bearing premise

The method assumes that the classifier's tentative pseudo-labels on unlabeled target images are accurate enough after a few training iterations that the PLMMD weighting aligns the same classes across domains instead of reinforcing the classifier's own mistakes.

Editorial extensions

If this is right

  • On Office-Home, IDAL reports an average accuracy of 73.5%, about 0.8 points above FixBi, the strongest prior CNN baseline in the comparison table.
  • On VisDA-2017, IDAL reports 87.4% average accuracy, about 0.2 points above FixBi, with notably larger per-class gains on classes such as truck and knife.
  • On Office-31 and DomainNet, the method is comparable rather than stronger: 90.8% versus FixBi's 91.4% on Office-31, and a 31.1% average on DomainNet with a ResNet-101 plus FPN.
  • The ablation study shows each added loss improves Office-Home average accuracy, from 65.8% for the CDAN base to 72.7% with MMD added, 73.2% with MMD and MCC, and 73.5% with PLMMD included.
  • The t-SNE visualizations support the convergence claim by showing target-domain classes separating into distinct clusters after only a few epochs of training.

Reading between the lines

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

  • If PLMMD's mechanism is real, the same class-conditional weighting should transfer to dense prediction tasks such as semantic segmentation, where per-pixel pseudo-labels could weight alignment more strongly for small or rare object classes; the paper lists this as future work.
  • A natural stress test the paper does not run is to corrupt pseudo-labels by shuffling them; if accuracy does not fall, the loss's benefit would have to come from the other regularization terms rather than from class-conditioned alignment.
  • The convergence-speed claim could be quantified by reporting accuracy at every epoch for IDAL versus FixBi, since the paper only states the effect and shows snapshots after 4 and 9 epochs; such curves would let others check whether faster convergence is steady or just a transient advantage.
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. This paper proposes IDAL, an unsupervised domain adaptation (UDA) method for image classification. The contributions are a feature extractor that combines ResNet with a feature pyramid network (FPN) and a combined loss consisting of a new pseudo-label maximum mean discrepancy (PLMMD) term together with information maximization, MMD, and MCC losses inside a CDAN-style adversarial framework. The authors report experiments on Office-Home, Office-31, VisDA-2017, and DomainNet, claiming state-of-the-art accuracy on the first three benchmarks and comparable performance on DomainNet, along with faster convergence.

Significance. If the reported results were supported by proper statistical evidence, the contribution would be a modest but useful empirical addition to UDA: a relatively simple architecture and loss combination that improves on a strong CDAN baseline on several benchmarks. The manuscript does not, however, provide code, multi-seed results, or a precise specification of the new loss, and the main accuracy claim is internally contradicted. As presented, the paper mainly demonstrates that the combination can be competitive, not that it surpasses the state of the art.

major comments (4)
  1. [Abstract, Section 4.3, Table 5] The abstract and introduction claim that IDAL 'generalizes better than state-of-the-art' on Office-31, but Section 4.3 describes the Office-31 result as 'comparable' and Table 5 reports IDAL at 90.8% average versus FixBi at 91.4%. These statements are contradictory, and the headline claim must be corrected or the experiment repeated before the paper can be accepted.
  2. [Tables 1, 4, and 5] The reported accuracy advantages over the strongest baseline are +0.8 points on Office-Home, +0.2 points on VisDA-2017, and -0.6 points on Office-31. All results are single runs with no seeds, standard deviations, or significance tests. With margins this small, the reader cannot determine whether the ordering reflects systematic improvement or run-to-run noise; the authors should report multiple seeds with confidence intervals, and the convergence-speed claim should be supported by learning curves rather than inferred from a single t-SNE statement.
  3. [Section 3.1, Eq. (9)] PLMMD is the claimed novelty, but the weight construction is described only in prose ('normalized dot products ... normalized by the count of common classes'), without an explicit formula. Because the weights depend on pseudo-labels produced by the model's own classifier, this loss is a self-training mechanism; without a precise weight definition and an analysis of pseudo-label accuracy, the method is not reproducible and the potential failure mode of reinforcing classifier errors is not addressed.
  4. [Section 4.3, Tables 2 and 3] For DomainNet the abstract promises 'comparable' performance, but IDAL's reported average of 31.1% in Table 3 is 2.2 points below MDD+SCDA (33.3%) and also below CDAN+SCDA (31.8%) in Table 2. If the comparison is not apples-to-apples, that should be stated explicitly; otherwise the claim of comparability needs to be weakened or substantiated with statistical evidence.
minor comments (5)
  1. [Abstract] There is a typo in the last sentence: 'comaparable' should be 'comparable'.
  2. [Section 2, first paragraph] The notation is inconsistent: the target domain is described with the symbol 'Ds', which is already used for the source domain; it should be 'Dt'.
  3. [Section 4.2 vs Section 4.3] The implementation details state that the backbone is ResNet-50, but Section 4.3 says ResNet-101 is used for DomainNet. This discrepancy should be clarified.
  4. [Equation (4)] The symbol T in Eq. (4) is used but never explicitly defined; the authors should state that it denotes the multilinear map introduced in the preceding paragraph.
  5. [Introduction and Related Work] The citation [25] appears to be incorrect: it is listed as a paper on web navigation path generation, but it is cited as a general GAN reference and as the source of synthetic-data generation in domain adaptation. A proper GAN reference should be supplied.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the paper's central claims are empirical benchmark results, and the proposed PLMMD loss is a training objective rather than a derivation that reduces to its own input.

full rationale

The paper does not derive its headline SOTA claim from any fitted value or from an equation defined in terms of the result. The reported improvements are empirical comparisons on Office-Home, Office-31, VisDA-2017, and DomainNet, supported by ablations in Table 6. The proposed PLMMD loss in Eq. (9) is a self-training-style objective whose weights are based on pseudo-labels from the model itself, but this is a training mechanism, not a circular derivation of the reported accuracy; any risk of pseudo-label error amplification is a correctness concern, not a logical circularity. The self-citation [10] by the same authors is used only to motivate the use of tailored loss combinations and is not load-bearing for the SOTA claim. No parameter is fitted to a subset and then renamed as a prediction, and no uniqueness theorem or prior result is invoked to force the architecture. The discrepancy between the abstract's Office-31 claim and Table 5 is a reporting/evidence issue, not circularity. Therefore the derivation chain is self-contained in the sense relevant to this analysis.

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

The paper's contribution reduces to (a) appending an FPN to ResNet-50 and (b) adding a class-weighted MMD variant, PLMMD, on top of CDAN, IM, MMD, and MCC. The new loss rests on an unvalidated pseudo-label quality assumption and its weight formula is under-specified. Four hyperparameters are tuned separately per dataset, which is a form of fitting to the reported benchmarks. No new physical or mathematical entities are postulated.

free parameters (4)
  • β (information maximization loss weight) = 0.05 for all four datasets
    Reported in Section 4.2; selected per dataset with no sensitivity analysis or cross-validation.
  • γ (minimum class confusion loss weight) = 0.1 (Office-31), 0.21 (Office-Home), 0.3 (VisDA-2017), 0.01 (DomainNet)
    Tuned separately per benchmark in Section 4.2; no study of robustness to these values.
  • δ (MMD loss weight) = 0.15 (Office-31), 0.25 (Office-Home), 0.25 (VisDA-2017), 0.2 (DomainNet)
    Same per-dataset tuning in Section 4.2; affects the balance between MMD and other losses.
  • η (PLMMD loss weight) = 0.15 (Office-31), 0.25 (Office-Home), 0.25 (VisDA-2017), 0.25 (DomainNet)
    Same per-dataset tuning in Section 4.2; directly controls the proposed loss's contribution.
assumptions (4)
  • standard math MMD with a characteristic kernel is a valid two-sample discrepancy, so minimizing MMD reduces the distribution gap (Eq. 8).
    Inherited from kernel two-sample testing via Long et al. [21]; the paper relies on this without re-deriving it.
  • domain assumption Source and target domains share the same label set, and all target classes appear in the source domain.
    PLMMD weights only classes 'common to both datasets' (Section 3.1); partial or open-set domain adaptation is not handled.
  • ad hoc to paper Target pseudo-labels become reliable after a few training iterations.
    Section 3.1 states pseudo-labels are generated 'after the first few training iterations' but provides no accuracy analysis, confidence threshold, or sensitivity study.
  • standard math CDAN's minimax adversarial training (Eq. 4) yields domain-invariant features when conditioned on classifier predictions.
    Adopted from Long et al. [22]; the paper builds on this framework without modifying or re-deriving its theory.

how reviews work

0 comments
Cite this review

Pith. "Pith review of IDAL: Improved Domain Adaptive Learning for Natural Images Dataset." pith.science (2026). https://pith.science/paper/SVN2VUKW

@misc{pith2026250617931,
  author       = {Pith},
  title        = {Pith review of: IDAL: Improved Domain Adaptive Learning for Natural Images Dataset},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/SVN2VUKW}},
  note         = {Machine review of arXiv:2506.17931}
}
read the original abstract

We present a novel approach for unsupervised domain adaptation (UDA) for natural images. A commonly-used objective for UDA schemes is to enhance domain alignment in representation space even if there is a domain shift in the input space. Existing adversarial domain adaptation methods may not effectively align different domains of multimodal distributions associated with classification problems. Our approach has two main features. Firstly, its neural architecture uses the deep structure of ResNet and the effective separation of scales of feature pyramidal network (FPN) to work with both content and style features. Secondly, it uses a combination of a novel loss function and judiciously selected existing loss functions to train the network architecture. This tailored combination is designed to address challenges inherent to natural images, such as scale, noise, and style shifts, that occur on top of a multi-modal (multi-class) distribution. The combined loss function not only enhances model accuracy and robustness on the target domain but also speeds up training convergence. Our proposed UDA scheme generalizes better than state-of-the-art for CNN-based methods on Office-Home, Office-31, and VisDA-2017 datasets and comaparable for DomainNet dataset.

Figures

Figures reproduced from arXiv: 2506.17931 by the authors.

Figure 1
Figure 1. The CDAN [22] architectural framework, illustrating the joint reduction of the cross-domain gap by the conditional domain discriminator D, with domain￾specific feature representation f and classifier prediction g at its core. The symbol ⊗ signifies a multilinear mapping operation. feature maps, has proven successful in embedding joint or conditional distribu￾tions into reproducing kernel Hilbert spaces [31]. In addi… view at source ↗
Figure 2
Figure 2. Two-dimensional visual representation, generated using t-SNE [ [PITH_FULL_IMAGE:figures/full_fig_p011_2.png] view at source ↗

Review history (2 revisions) →

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

38 extracted references · 27 canonical work pages

  1. [1]

    f-domain adver- sarial learning: Theory and algorithms

    David Acuna, Guojun Zhang, Marc T Law, and Sanja Fidler. f-domain adver- sarial learning: Theory and algorithms. InInternational Conference on Machine Learning, pages 66–75. PMLR, 2021

    work page 2021

  2. [2]

    A theory of learning from different domains

    Shai Ben-David, John Blitzer, Koby Crammer, Alex Kulesza, Fernando Pereira, and Jennifer Wortman Vaughan. A theory of learning from different domains. Machine learning, 79:151–175, 2010

    2010

  3. [3]

    Progressive feature alignment for unsupervised domain adaptation

    Chaoqi Chen, Weiping Xie, Wenbing Huang, Yu Rong, Xinghao Ding, Yue Huang, Tingyang Xu, and Junzhou Huang. Progressive feature alignment for unsupervised domain adaptation. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 627–636, 2019

    work page 2019

  4. [4]

    Towards discriminability and diversity: Batch nuclear-norm maximization under label insufficient situations

    Shuhao Cui, Shuhui Wang, Junbao Zhuo, Liang Li, Qingming Huang, and Qi Tian. Towards discriminability and diversity: Batch nuclear-norm maximization under label insufficient situations. InProceedings of the IEEE/CVF conference on com- puter vision and pattern recognition, pages 3941–3950, 2020

    work page 2020

  5. [5]

    Graduallyvanishingbridgeforadversarialdomainadaptation

    Shuhao Cui, Shuhui Wang, Junbao Zhuo, Chi Su, Qingming Huang, and Qi Tian. Graduallyvanishingbridgeforadversarialdomainadaptation. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 12455– 12464, 2020

    work page 2020

  6. [6]

    An image is worth 16x16 words: Transformers for image recognition at scale.arXiv preprint arXiv:2010.11929, 2020

    Alexey Dosovitskiy, Lucas Beyer, Alexander Kolesnikov, Dirk Weissenborn, Xi- aohua Zhai, Thomas Unterthiner, Mostafa Dehghani, Matthias Minderer, Georg Heigold, Sylvain Gelly, et al. An image is worth 16x16 words: Transformers for image recognition at scale.arXiv preprint arXiv:2010.11929, 2020

    arXiv 2010

  7. [7]

    Unsupervised domain adaptation by back- propagation

    Yaroslav Ganin and Victor Lempitsky. Unsupervised domain adaptation by back- propagation. InInternational conference on machine learning, pages 1180–1189. PMLR, 2015. 14 Ravi Kant Gupta, Shounak Das, and Amit Sethi

    work page 2015

  8. [8]

    Domain- adversarial training of neural networks.The journal of machine learning research, 17(1):2096–2030, 2016

    Yaroslav Ganin, Evgeniya Ustinova, Hana Ajakan, Pascal Germain, Hugo Larochelle, François Laviolette, Mario Marchand, and Victor Lempitsky. Domain- adversarial training of neural networks.The journal of machine learning research, 17(1):2096–2030, 2016

    work page 2016

Show all 38 references
  1. [9]

    Robertson

    Jian Gao, Yang Hua, Guosheng Hu, Chi Wang, and Neil M. Robertson. Reducing distributional uncertainty by mutual information maximisation and transferable feature learning. InComputer Vision – ECCV 2020: 16th European Conference, Glasgow, UK, August 23–28, 2020, Proceedings, Pa...

    2020

  2. [10]

    Domain-adaptive learning: Unsu- pervised adaptation for histology images with improved loss function combination

    Ravi Kant Gupta, Shounak Das, and Amit Sethi. Domain-adaptive learning: Unsu- pervised adaptation for histology images with improved loss function combination. arXiv preprint arXiv:2309.17172, 2023

    2023 arXiv

  3. [11]

    Deep residual learning for image recognition

    Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. Deep residual learning for image recognition. InProceedings of the IEEE conference on computer vision and pattern recognition, pages 770–778, 2016

    2016

  4. [12]

    Deep residual learning for image recognition

    S Jian, H Kaiming, R Shaoqing, and Z Xiangyu. Deep residual learning for image recognition. InIEEE Conference on Computer Vision & Pattern Recognition, pages 770–778, 2016

    2016

  5. [13]

    Minimum class confusion for versatile domain adaptation

    Ying Jin, Ximei Wang, Mingsheng Long, and Jianmin Wang. Minimum class confusion for versatile domain adaptation. InComputer Vision–ECCV 2020: 16th European Conference, Glasgow, UK, August 23–28, 2020, Proceedings, Part XXI 16, pages 464–480. Springer, 2020

    2020

  6. [14]

    Deep adversarial atten- tion alignment for unsupervised domain adaptation: the benefit of target expecta- tion maximization

    Guoliang Kang, Liang Zheng, Yan Yan, and Yi Yang. Deep adversarial atten- tion alignment for unsupervised domain adaptation: the benefit of target expecta- tion maximization. InProceedings of the European conference on computer vision (ECCV), pages 401–416, 2018

    2018

  7. [15]

    Discriminative clustering by regularized information maximization.Advances in neural information processing systems, 23, 2010

    Andreas Krause, Pietro Perona, and Ryan Gomes. Discriminative clustering by regularized information maximization.Advances in neural information processing systems, 23, 2010

    2010

  8. [16]

    Attending to discriminative certainty for domain adaptation

    Vinod Kumar Kurmi, Shanu Kumar, and Vinay P Namboodiri. Attending to discriminative certainty for domain adaptation. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 491–500, 2019

    2019

  9. [17]

    Sliced wasserstein discrepancy for unsupervised domain adaptation

    Chen-Yu Lee, Tanmay Batra, Mohammad Haris Baig, and Daniel Ulbricht. Sliced wasserstein discrepancy for unsupervised domain adaptation. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 10285– 10295, 2019

    2019

  10. [18]

    Semantic concentration for domain adaptation.CoRR, abs/2108.05720, 2021

    Shuang Li, Mixue Xie, Fangrui Lv, Chi Harold Liu, Jian Liang, Chen Qin, and Wei Li. Semantic concentration for domain adaptation.CoRR, abs/2108.05720, 2021

    2021 arXiv

  11. [19]

    Do we really need to access the source data? source hypothesis transfer for unsupervised domain adaptation

    Jian Liang, Dapeng Hu, and Jiashi Feng. Do we really need to access the source data? source hypothesis transfer for unsupervised domain adaptation. InInterna- tional conference on machine learning, pages 6028–6039. PMLR, 2020

    2020

  12. [20]

    Feature pyramid networks for object detection

    Tsung-Yi Lin, Piotr Dollár, Ross Girshick, Kaiming He, Bharath Hariharan, and Serge Belongie. Feature pyramid networks for object detection. InProceedings of the IEEE conference on computer vision and pattern recognition, pages 2117–2125, 2017

    2017

  13. [21]

    Learning trans- ferable features with deep adaptation networks

    Mingsheng Long, Yue Cao, Jianmin Wang, and Michael Jordan. Learning trans- ferable features with deep adaptation networks. InInternational conference on machine learning, pages 97–105. PMLR, 2015

    2015

  14. [22]

    Conditional adversarial domain adaptation.Advances in neural information processing systems, 31, 2018

    Mingsheng Long, Zhangjie Cao, Jianmin Wang, and Michael I Jordan. Conditional adversarial domain adaptation.Advances in neural information processing systems, 31, 2018. IDAL: Improved Domain Adaptive Learning for Natural Images Dataset 15

    2018

  15. [23]

    Decoupled weight decay regularization

    Ilya Loshchilov and Frank Hutter. Decoupled weight decay regularization. In International Conference on Learning Representations, 2019

    2019

  16. [24]

    Fixbi: Bridg- ing domain spaces for unsupervised domain adaptation

    Jaemin Na, Heechul Jung, Hyung Jin Chang, and Wonjun Hwang. Fixbi: Bridg- ing domain spaces for unsupervised domain adaptation. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 1094– 1103, 2021

    2021

  17. [25]

    Generation of realistic navigation paths for web site testing using rnn and gan.Journal of Web Engineering, pages 2571–2604, 2021

    Silvio Pavanetto and Marco Brambilla. Generation of realistic navigation paths for web site testing using rnn and gan.Journal of Web Engineering, pages 2571–2604, 2021

    2021

  18. [26]

    Multi-adversarial domain adaptation

    Zhongyi Pei, Zhangjie Cao, Mingsheng Long, and Jianmin Wang. Multi-adversarial domain adaptation. InProceedings of the AAAI conference on artificial intelli- gence, volume 32, 2018

    2018

  19. [27]

    Moment matching for multi-source domain adaptation

    Xingchao Peng, Qinxun Bai, Xide Xia, Zijun Huang, Kate Saenko, and Bo Wang. Moment matching for multi-source domain adaptation. InProceedings of the IEEE/CVF international conference on computer vision, pages 1406–1415, 2019

    2019

  20. [28]

    A closer look at smoothness in domain adversar- ial training

    Harsh Rangwani, Sumukh K Aithal, Mayank Mishra, Arihant Jain, and Venkatesh Babu Radhakrishnan. A closer look at smoothness in domain adversar- ial training. InInternational Conference on Machine Learning, pages 18378–18399. PMLR, 2022

    2022

  21. [29]

    Adapting visual cat- egory models to new domains

    Kate Saenko, Brian Kulis, Mario Fritz, and Trevor Darrell. Adapting visual cat- egory models to new domains. InComputer Vision–ECCV 2010: 11th European Conference on Computer Vision, Heraklion, Crete, Greece, September 5-11, 2010, Proceedings, Part IV 11, pages 213–226. Springer, 2010

    2010

  22. [30]

    Maxi- mum classifier discrepancy for unsupervised domain adaptation

    Kuniaki Saito, Kohei Watanabe, Yoshitaka Ushiku, and Tatsuya Harada. Maxi- mum classifier discrepancy for unsupervised domain adaptation. InProceedings of the IEEE conference on computer vision and pattern recognition, pages 3723–3732, 2018

    2018

  23. [31]

    Hilbert space em- beddings of conditional distributions with applications to dynamical systems

    Le Song, Jonathan Huang, Alex Smola, and Kenji Fukumizu. Hilbert space em- beddings of conditional distributions with applications to dynamical systems. In Proceedings of the 26th Annual International Conference on Machine Learning, pages 961–968, 2009

    2009

  24. [32]

    Unsupervised domain adaptation via struc- turally regularized deep clustering

    Hui Tang, Ke Chen, and Kui Jia. Unsupervised domain adaptation via struc- turally regularized deep clustering. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 8725–8735, 2020

    2020

  25. [33]

    Patches are all you need?arXiv preprint arXiv:2201.09792, 2022

    Asher Trockman and J Zico Kolter. Patches are all you need?arXiv preprint arXiv:2201.09792, 2022

    2022 arXiv

  26. [34]

    Adversarial discrim- inative domain adaptation.CoRR, abs/1702.05464, 2017

    Eric Tzeng, Judy Hoffman, Kate Saenko, and Trevor Darrell. Adversarial discrim- inative domain adaptation.CoRR, abs/1702.05464, 2017

    2017 arXiv

  27. [35]

    Visualizing data using t-sne.Jour- nal of machine learning research, 9(11), 2008

    Laurens Van der Maaten and Geoffrey Hinton. Visualizing data using t-sne.Jour- nal of machine learning research, 9(11), 2008

    2008

  28. [36]

    Deep hashing network for unsupervised domain adaptation

    Hemanth Venkateswara, Jose Eusebio, Shayok Chakraborty, and Sethuraman Pan- chanathan. Deep hashing network for unsupervised domain adaptation. InPro- ceedings of the IEEE conference on computer vision and pattern recognition, pages 5018–5027, 2017

    2017

  29. [37]

    How transferable are features in deep neural networks?Advances in neural information processing systems, 27, 2014

    Jason Yosinski, Jeff Clune, Yoshua Bengio, and Hod Lipson. How transferable are features in deep neural networks?Advances in neural information processing systems, 27, 2014

    2014

  30. [38]

    Bridging the- ory and algorithm for domain adaptation

    Yuchen Zhang, Tianle Liu, Mingsheng Long, and Michael Jordan. Bridging the- ory and algorithm for domain adaptation. In Kamalika Chaudhuri and Ruslan Salakhutdinov, editors,Proceedings of the 36th International Conference on Ma- chine Learning, volume 97 ofProceedings of Machi...

    2019

Pith tools

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