Pith. sign in

REVIEW 3 major objections 4 minor 56 references

Improving Out-of-Domain Robustness with Targeted Augmentation in Frequency and Pixel Spaces

T0 review · 3 major / 4 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 A single augmentation method that mixes frequency amplitudes and pixel content improves out-of-domain accuracy on four real-world benchmarks by up to 9.1 percentage points.

desk verdict Useful combination of known augmentation tricks, with a real protocol problem on the Galaxy10 benchmark that should be fixed before the +9.1% claim is taken at face value. read the letter →

arxiv 2505.12317 v1 pith:2R5LS4IX submitted 2025-05-18 cs.CV cs.LG

Ruoqi Wang , Haitao Wang , Shaojie Guo , Qiong Luo This is my paper · ORCID
classification cs.CVcs.LG
keywords Frequency-PixelConnectout-of-domainrobustnessdomainadaptationfrequency-spaceaugmentationpixel-spacemixingFourieramplitudeinterpolationconnectivityanalysistargeted
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

The paper claims that one dataset-agnostic augmentation scheme can improve out-of-domain robustness across four qualitatively different real-world domains, replacing hand-designed, dataset-specific augmentations. The scheme mixes the Fourier amplitude spectrum of a source image with that of a target image while keeping the source phase, then blends the result with a pixel-space mix; this perturbs domain-specific appearance features without destroying label-relevant structure. On wildlife photos, histopathology slides, bird audio spectrograms, and galaxy images, the method reports OOD gains of +3.0, +4.9, +6.4, and +9.1 percentage points over generic baselines, and also surpasses augmentations designed specifically for each dataset. If the claim is right, practitioners can get large unsupervised domain-adaptation gains without per-dataset expert analysis.

What carries the argument

The central mechanism is Frequency-Pixel Mixing, built on the Fourier amplitude/phase decomposition: amplitude carries low-level appearance and domain style, while phase carries semantic structure. The augmentation interpolates amplitude spectra within a random square crop, keeps the source phase, then fuses the frequency-augmented image with a pixel-space blend through a second mixing ratio $\lambda_2$. A pretrain-then-fine-tune protocol (linear probing then fine-tuning) carries the augmented data into the encoder, and connectivity is the diagnostic: binary classifiers trained to distinguish class-domain pairs produce ratios $\alpha/\gamma$ (same class, different domain over different class, different domain) and $\beta/\gamma$, which the paper uses to claim that its augmentation aligns same-class pairs across domains more strongly than it conflates different classes.

What would settle it

On Galaxy10, retrain with mixing targets drawn only from labeled DECaLS images or from a held-out unlabeled pool that is never used for evaluation, and compare out-of-domain accuracy on the SDSS subset against RandAugment. If the improvement largely disappears, the reported +9.1 percent gain depends on test-set mixing rather than on the augmentation mechanism itself.

Watch

Extended reading notes

Core claim

Frequency-Pixel Connect is a two-stage domain-adaptation framework: pretrain with generic augmentations, then fine-tune with a targeted augmentation that operates in both frequency and pixel spaces. For a source image $x_1$ and a target-domain image $x_2$, it linearly interpolates their amplitude spectra inside a randomly cropped square region, keeps the source phase, inverts the Fourier transform, and blends this frequency-augmented image with a pixel-wise mix of $x_1$ and $x_2$. The resulting augmented sample preserves the semantic content of the source through the phase and a tunable fusion ratio while injecting target-domain appearance statistics through amplitude and pixels. The paper argues, using connectivity ratios, that this increases same-class cross-domain connectivity relative to different-class cross-domain connectivity, aligning representations across domains without collapsing class structure.

Load-bearing premise

The central claim depends on treating the unlabeled target data as fair game for training-time mixing; on Galaxy10 those mixing targets come from the evaluation test set, so the +9.1 percent gain is evidence of generalization only if using the test set this way is not considered leakage.

Editorial extensions

If this is right

  • One generic augmentation recipe can serve as a drop-in fine-tuning module for unsupervised domain adaptation with unlabeled target data, removing the need for dataset-specific augmentations such as Copy-Paste or Stain Color Jitter.
  • On all four benchmarks, the method reaches out-of-domain performance above every generic baseline tested and above the dataset-specific targeted-augmentation baselines tested.
  • Frequency-space perturbation is the larger contributor to the out-of-domain gains, while pixel-space blending adds complementary detail, according to the paper's ablation study.
  • The connectivity ratio $\alpha/\gamma$ is the proposed explanation for the gains: raising same-class cross-domain connectivity more than different-class cross-domain connectivity should track better out-of-domain accuracy, and the paper reports that its augmentation does this.

Reading between the lines

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

  • A conservative reading of the Galaxy10 result treats it as an upper bound for transductive domain adaptation rather than proof of generalization to a fully unseen telescope, because the paper's Appendix C.1 states that mixing targets are sampled from the unlabeled test set when the training data has a single source domain.
  • The amplitude/phase split suggests a diagnostic: measuring the Fourier amplitude distance between source and target domains should predict which benchmarks benefit most from frequency-space mixing, with texture-level shifts favoring pixel-space blending.
  • The two mixing ratios $\lambda_1$ and $\lambda_2$ are the main sensitivity, so an automatic rule that adapts them from validation connectivity could remove the manual tuning the paper itself lists as a limitation.
  • Because the method already treats bird audio spectrograms as images, it could transfer to other Fourier-factorizable signals such as medical volumes or radar data without new design.
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

3 major / 4 minor

Summary. The manuscript proposes Frequency-Pixel Connect, an unsupervised domain adaptation method that augments source-domain images by mixing their Fourier amplitude spectra with target-domain images (Eq. 4) and by pixel-space blending (Eq. 7), then fusing the two augmented views with a second mixing ratio (Eq. 8). The method is evaluated under a contrastive-pretraining plus LP-FT protocol on four benchmarks: iWildCam, Camelyon17, BirdCalls, and Galaxy10, with reported OOD gains of +3.0% F1, +4.9% accuracy, +6.4% F1, and +9.1% accuracy. The authors also present an ablation study, a connectivity analysis based on the ratios alpha/gamma and beta/gamma, and a sensitivity study of the two mixing parameters.

Significance. If the reported four-benchmark results were valid, the paper would provide strong evidence that a dataset-agnostic augmentation acting in both frequency and pixel spaces can match or exceed dataset-specific targeted augmentations. Strengths include head-to-head comparisons with Connect Later and with dataset-specific methods such as Copy-Paste and Stain Color Jitter, ablations separating the frequency and pixel contributions, five-seed averaging in the experimental protocol, and a mechanistic connectivity analysis. The Galaxy10 protocol in Appendix C.1 is a serious flaw: test-set images are used as mixing targets during training and the same test set is then used for evaluation, so the +9.1% Galaxy10 gain cannot be interpreted as out-of-domain generalization. The remaining three benchmarks are less affected, so the central methodological idea remains plausible, but the headline claim of consistent gains across four diverse benchmarks needs revision.

major comments (3)
  1. [Appendix C.1, Galaxy10 protocol] The text states that 'since the training data only contains images from a single telescope, we randomly sample images from the unlabeled test set as mixing targets.' The same SDSS set is then used to compute the +9.1% OOD accuracy reported in Section 4.2. Equations (4) and (7) place the pixel content and amplitude statistics of test images directly into the training distribution, and none of the comparison baselines receive target-domain images in this way. This is not a clean evaluation of generalization to an unseen target domain; it is direct exposure to the evaluation distribution. The four-benchmark claim is therefore not supported, and the Galaxy10 result should be either re-run with target mixing images disjoint from the test set or removed from the headline claims.
  2. [Section 4.2, Table 2 and Figure 4] Although Appendix C.1 states that results are averaged over five random seeds, no standard deviations or confidence intervals are reported for any benchmark. Without this information the claims of 'significant' and 'consistent' improvements over the strongest baselines cannot be assessed, especially for the smaller gaps such as the iWildCam OOD F1 difference and the BirdCalls OOD F1 difference. Please report variance across seeds for the main tables and figures.
  3. [Section 4.3, Table 3] The connectivity analysis is presented as an empirical evaluation, but the augmentation was explicitly designed in Section 2.4 to increase alpha/gamma by randomizing spurious domain-dependent features while preserving label-relevant semantics. Table 3 therefore largely confirms the design objective rather than providing independent evidence that connectivity explains the OOD gains. The claim that Frequency-Pixel Connect 'significantly improves cross-domain connectivity' should be repositioned as a design sanity check, with the OOD accuracy results carrying the primary evidential weight.
minor comments (4)
  1. [Appendix C.1, Galaxy10] The 'predefined label conversion rule' between Galaxy10 DECaLS and SDSS labels is referenced but never specified; please provide the rule or a pointer to the code so that the evaluation is reproducible.
  2. [Appendix C.1 and experimental paragraphs] The word 'Othor' appears in the iWildCam, Camelyon17, and BirdCalls implementation paragraphs; it should be 'Other'.
  3. [References] References [16] and [17] are duplicate entries for the same CyCADA paper; one should be removed and the in-text citations adjusted.
  4. [Table 3] The connectivity values are reported as single point estimates without any measure of variability, even though they are obtained by training binary classifiers; please report variance across seeds or otherwise describe the precision of these estimates.

Circularity Check

1 steps flagged · score 6.0 of 10

Galaxy10 OOD gain is contaminated by using the evaluation test set as the augmentation source; other benchmarks remain independent.

  1. fitted input called prediction [Appendix C.1 (Galaxy10), Section 4.2, Eq. (4), Eq. (7)]
    "For Galaxy10, since the training data only contains images from a single telescope, we randomly sample images from the unlabeled test set as mixing targets to generate cross-domain augmented images."

    Eq. (7) defines the pixel-space augmented image as x_hat_p = (1 - lambda_1) x1 + lambda_1 x2, and Eq. (4) defines the frequency-space augmented amplitude as A(x_hat) = (1 - lambda) A(x1) + lambda A(x2). Appendix C.1 sets x2 for Galaxy10 to be images sampled from the unlabeled test set, i.e., from the same SDSS images that Section 4.2 scores to report 62.3% OOD accuracy (+9.1% over RandAugment). The fine-tuning inputs therefore contain the exact pixel content and amplitude spectra of the evaluation images, so the reported Galaxy10 gain is not a measure of generalization to an unseen target domain; it is partly forced by direct exposure to the test set. Generic baselines such as RandAugment, MixUp, and CutMix receive no such target-test content, making the comparison asymmetric.

full rationale

Frequency-Pixel Connect is a constructive augmentation method, not a fitted model of the benchmarks, and its headline gains on iWildCam, Camelyon17, and BirdCalls are measured against external WILDS/LifeCLEF test sets with target-domain images taken from training splits. I did not find a load-bearing self-citation chain: the cited connectivity result [36] and targeted-augmentation baselines [9,30] are external prior work, not the authors' own, and the amplitude/phase mixing is justified by classical results plus the paper's controlled experiments. The connectivity analysis in Tables 1 and 3 is a diagnostic of the augmentation's intended effect rather than the source of the headline OOD numbers, and the paper's acknowledged dependence on tuning lambda_1 and lambda_2 is hyperparameter sensitivity, not circularity. The only load-bearing circular step is the Galaxy10 protocol: Appendix C.1 uses the unlabeled SDSS test set as the source of mixing targets for Eq. (7) and Eq. (4), and the same SDSS set is scored in Section 4.2 to report the +9.1% OOD gain. That makes the Galaxy10 component of the 'four benchmarks' claim a transductive-fit artifact rather than an OOD generalization result, while the other three benchmarks retain independent content.

Assumptions & free parameters 5 free parameters · 4 assumptions · 0 invented entities

The central claim rests on the heuristic that amplitude/phase decomposition separates style from semantics, and on the connectivity heuristic of Shen et al. The main free parameters are the mixing intensities, which are tuned per dataset and not fully reported.

free parameters (5)
  • lambda_1 (pixel-space blending ratio)
    Eq. 7 controls the degree of pixel blending; Figure 5 shows OOD performance varies strongly with it, but final values per dataset are not reported.
  • lambda_2 (fusion ratio)
    Eq. 8 combines the frequency-augmented and pixel-augmented images; it is tuned on OOD validation but reported values are absent.
  • eta (upper bound of lambda in frequency mixing)
    In Eq. 4, lambda is sampled from U(0, eta). For Camelyon17, augmentation strength is sampled from [0.05,0.1]; other datasets give no value for eta.
  • r (crop ratio in frequency amplitude interpolation)
    The text says a square crop of ratio r is used, but the value of r is never given for any experiment.
  • augmentation probability
    Sampled from Uniform[0.5,0.9] per dataset and selected by OOD validation, so it is a tuned hyperparameter.
assumptions (4)
  • domain assumption Input features decompose into label/domain-dependent and label/domain-independent components (x_obj, x_d:robust, x_d:spu, x_noise) with the stated (in)dependences.
    Section 2.1 adopts this framework from [9,36]; the paper's motivation and post-hoc explanations rest on it.
  • domain assumption Phase carries semantic structure and amplitude carries style/appearance, so amplitude mixing preserves label-relevant content.
    Section 2.4 and Eqs. 4-5 rely on this, citing [46] and classical phase studies.
  • domain assumption Connectivity ratios alpha/gamma and beta/gamma estimate target accuracy via target accuracy approximately (alpha/gamma)^w1 (beta/gamma)^w2 with w1 > w2.
    Section 4.3 uses this from [36] to interpret Table 3; it is a heuristic, not a theorem.
  • domain assumption Generic contrastive pretraining (SwAV) plus linear-probing-then-fine-tuning provides a sound base for OOD robustness.
    Used as the base pipeline from [30]; the paper does not derive or justify it beyond following prior work.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Improving Out-of-Domain Robustness with Targeted Augmentation in Frequency and Pixel Spaces." pith.science (2026). https://pith.science/paper/2R5LS4IX

@misc{pith2026250512317,
  author       = {Pith},
  title        = {Pith review of: Improving Out-of-Domain Robustness with Targeted Augmentation in Frequency and Pixel Spaces},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/2R5LS4IX}},
  note         = {Machine review of arXiv:2505.12317}
}
read the original abstract

Out-of-domain (OOD) robustness under domain adaptation settings, where labeled source data and unlabeled target data come from different distributions, is a key challenge in real-world applications. A common approach to improving OOD robustness is through data augmentations. However, in real-world scenarios, models trained with generic augmentations can only improve marginally when generalized under distribution shifts toward unlabeled target domains. While dataset-specific targeted augmentations can address this issue, they typically require expert knowledge and extensive prior data analysis to identify the nature of the datasets and domain shift. To address these challenges, we propose Frequency-Pixel Connect, a domain-adaptation framework that enhances OOD robustness by introducing a targeted augmentation in both the frequency space and pixel space. Specifically, we mix the amplitude spectrum and pixel content of a source image and a target image to generate augmented samples that introduce domain diversity while preserving the semantic structure of the source image. Unlike previous targeted augmentation methods that are both dataset-specific and limited to the pixel space, Frequency-Pixel Connect is dataset-agnostic, enabling broader and more flexible applicability beyond natural image datasets. We further analyze the effectiveness of Frequency-Pixel Connect by evaluating the performance of our method connecting same-class cross-domain samples while separating different-class examples. We demonstrate that Frequency-Pixel Connect significantly improves cross-domain connectivity and outperforms previous generic methods on four diverse real-world benchmarks across vision, medical, audio, and astronomical domains, and it also outperforms other dataset-specific targeted augmentation methods.

Figures

Figures reproduced from arXiv: 2505.12317 by the authors.

Figure 1
Figure 1. Feature decomposition and augmentation examples. Top: For each dataset (iWildCam, Camelyon17, BirdCalls, Galaxy10), we show a source image, its corresponding augmented image generated by our method. Bottom: Based on the decomposition framework [36], we annotate representative features in the datasets across xobj, xd:robust, xd:spu, xnoise. Our method effectively randomizes xd:spu, varies xd:robust while preserving o… view at source ↗
Figure 2
Figure 2. The figure shows the amplitude-only and phase-only reconstructions of the two input images, and the process where the amplitudes of Image 1 and Image 2 are first mixed, and the mixed amplitude is combined with the phase of Image 2 to generate an augmentation of Image 2. Our method is motivated by observing and analyzing how frequency and pixel-space mixing affect different feature components under domain shift. Out-… view at source ↗
Figure 3
Figure 3. Overview of our Frequency-Pixel Mixing strategy. Given two images, we first do amplitude [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: We plot the in-domain (ID) performance versus out-of-domain (OOD) performance [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]
Figure 5
Figure 5. Figure 5: Effect of Mixing Ratios on ID and OOD Performance. [PITH_FULL_IMAGE:figures/full_fig_p009_5.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

56 extracted references · 33 canonical work pages

  1. [1]

    Peter Bandi, Oscar Geessink, Quirine Manson, Marcory Van Dijk, Maschenka Balkenhol, Meyke Hermsen, Babak Ehteshami Bejnordi, Byungjae Lee, Kyunghyun Paeng, Aoxiao Zhong, et al. From detection of individual metastases to classification of lymph node status at the patient level: the camelyon17 challenge.IEEE transactions on medical imaging, 38(2):550–560, 2018

    2018

  2. [2]

    The iwildcam 2021 competi- tion dataset.arXiv preprint arXiv:2105.03494, 2021

    Sara Beery, Arushi Agarwal, Elijah Cole, and Vighnesh Birodkar. The iwildcam 2021 competi- tion dataset.arXiv preprint arXiv:2105.03494, 2021

    arXiv 2021

  3. [3]

    Recognition in terra incognita

    Sara Beery, Grant Van Horn, and Pietro Perona. Recognition in terra incognita. InProceedings of the European conference on computer vision (ECCV), pages 456–473, 2018

    2018

  4. [4]

    Unsupervised learning of visual features by contrasting cluster assignments.Advances in neural information processing systems, 33:9912–9924, 2020

    Mathilde Caron, Ishan Misra, Julien Mairal, Priya Goyal, Piotr Bojanowski, and Armand Joulin. Unsupervised learning of visual features by contrasting cluster assignments.Advances in neural information processing systems, 33:9912–9924, 2020

    2020

  5. [5]

    A simple framework for contrastive learning of visual representations

    Ting Chen, Simon Kornblith, Mohammad Norouzi, and Geoffrey Hinton. A simple framework for contrastive learning of visual representations. InInternational conference on machine learning, pages 1597–1607. PmLR, 2020

    2020

  6. [6]

    Randaugment: Practical automated data augmentation with a reduced search space

    Ekin D Cubuk, Barret Zoph, Jonathon Shlens, and Quoc V Le. Randaugment: Practical automated data augmentation with a reduced search space. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition workshops, pages 702–703, 2020

    2020

  7. [7]

    Improved regularization of convolutional neural networks with cutout.arXiv preprint arXiv:1708.04552, 2017

    Terrance DeVries and Graham W Taylor. Improved regularization of convolutional neural networks with cutout.arXiv preprint arXiv:1708.04552, 2017

    arXiv 2017

  8. [8]

    Domain-adversarial training of neural networks

    Yaroslav Ganin, Evgeniya Ustinova, Hana Ajakan, Pascal Germain, Hugo Larochelle, François Laviolette, Mario March, and Victor Lempitsky. Domain-adversarial training of neural networks. Journal of machine learning research, 17(59):1–35, 2016

    2016

Show all 56 references
  1. [9]

    Out-of-domain robustness via targeted augmentations

    Irena Gao, Shiori Sagawa, Pang Wei Koh, Tatsunori Hashimoto, and Percy Liang. Out-of-domain robustness via targeted augmentations. InInternational Conference on Machine Learning, pages 10800–10834. PMLR, 2023

    2023

  2. [10]

    In search of lost domain generalization

    Ishaan Gulrajani and David Lopez-Paz. In search of lost domain generalization. InInternational Conference on Learning Representations, 2021

    2021

  3. [11]

    Structural sparseness and spatial phase alignment in natural scenes.Journal of the Optical Society of America A, 24(7):1873–1885, 2007

    Bruce C Hansen and Robert F Hess. Structural sparseness and spatial phase alignment in natural scenes.Journal of the Optical Society of America A, 24(7):1873–1885, 2007

    2007

  4. [12]

    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

  5. [13]

    The many faces of robustness: A critical analysis of out-of-distribution generalization

    Dan Hendrycks, Steven Basart, Norman Mu, Saurav Kadavath, Frank Wang, Evan Dorundo, Rahul Desai, Tyler Zhu, Samyak Parajuli, Mike Guo, et al. The many faces of robustness: A critical analysis of out-of-distribution generalization. InProceedings of the IEEE/CVF international co...

    2021

  6. [14]

    Augmix: A simple data processing method to improve robustness and uncertainty

    Dan Hendrycks, Norman Mu, Ekin Dogus Cubuk, Barret Zoph, Justin Gilmer, and Balaji Lakshminarayanan. Augmix: A simple data processing method to improve robustness and uncertainty. InInternational Conference on Learning Representations, 2020

    2020

  7. [15]

    Galaxy10 decals dataset

    Leung Henry. Galaxy10 decals dataset. https://github.com/henrysky/Galaxy10, 2021

    2021

  8. [16]

    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. InInternational conference on machine learning, pages 1989–1998. Pmlr, 2018

    1989

  9. [17]

    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. InInternational conference on machine learning, pages 1989–1998. Pmlr, 2018. 10

    1989

  10. [18]

    Selecting data augmentation for simulating interventions

    Maximilian Ilse, Jakub M Tomczak, and Patrick Forré. Selecting data augmentation for simulating interventions. InInternational conference on machine learning, pages 4555–4562. PMLR, 2021

    2021

  11. [19]

    Overview of lifeclef 2022: an evaluation of machine-learning based species identification and species distribution prediction

    Alexis Joly, Hervé Goëau, Stefan Kahl, Lukáš Picek, Titouan Lorieul, Elijah Cole, Benjamin Deneu, Maximilien Servajean, Andrew Durso, Hervé Glotin, et al. Overview of lifeclef 2022: an evaluation of machine-learning based species identification and species distribution predict...

    2022

  12. [20]

    Contrastive adaptation network for unsupervised domain adaptation

    Guoliang Kang, Lu Jiang, Yi Yang, and Alexander G Hauptmann. Contrastive adaptation network for unsupervised domain adaptation. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 4893–4902, 2019

    2019

  13. [21]

    Wilds: A benchmark of in-the-wild distribution shifts

    Pang Wei Koh, Shiori Sagawa, Henrik Marklund, Sang Michael Xie, Marvin Zhang, Akshay Balsubramani, Weihua Hu, Michihiro Yasunaga, Richard Lanas Phillips, Irena Gao, et al. Wilds: A benchmark of in-the-wild distribution shifts. InInternational conference on machine learning, pa...

    2021

  14. [22]

    Imagenet classification with deep convolutional neural networks.Advances in neural information processing systems, 25, 2012

    Alex Krizhevsky, Ilya Sutskever, and Geoffrey E Hinton. Imagenet classification with deep convolutional neural networks.Advances in neural information processing systems, 25, 2012

    2012

  15. [23]

    Fine-tuning can distort pretrained features and underperform out-of-distribution

    Ananya Kumar, Aditi Raghunathan, Robbie Jones, Tengyu Ma, and Percy Liang. Fine-tuning can distort pretrained features and underperform out-of-distribution. InInternational Conference on Learning Representations, 2022

    2022

  16. [24]

    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

    2018

  17. [25]

    Ssmba: Self-supervised manifold based data augmentation for improving out-of-domain robustness.arXiv preprint arXiv:2009.10195, 2020

    Nathan Ng, Kyunghyun Cho, and Marzyeh Ghassemi. Ssmba: Self-supervised manifold based data augmentation for improving out-of-domain robustness.arXiv preprint arXiv:2009.10195, 2020

    2009 arXiv

  18. [26]

    Phase in speech and pictures

    A Oppenheim, Jae Lim, Gary Kopec, and SC Pohlig. Phase in speech and pictures. In ICASSP’79. IEEE International Conference on Acoustics, Speech, and Signal Processing, volume 4, pages 632–637. IEEE, 1979

    1979

  19. [27]

    The importance of phase in signals.Proceedings of the IEEE, 69(5):529–541, 1981

    Alan V Oppenheim and Jae S Lim. The importance of phase in signals.Proceedings of the IEEE, 69(5):529–541, 1981

    1981

  20. [28]

    Specaugment: A simple data augmentation method for automatic speech recognition.arXiv preprint arXiv:1904.08779, 2019

    Daniel S Park, William Chan, Yu Zhang, Chung-Cheng Chiu, Barret Zoph, Ekin D Cubuk, and Quoc V Le. Specaugment: A simple data augmentation method for automatic speech recognition.arXiv preprint arXiv:1904.08779, 2019

    1904 arXiv

  21. [29]

    A demonstration of the visual importance and flexibility of spatial-frequency amplitude and phase.Perception, 11(3):337–346, 1982

    Leon N Piotrowski and Fergus W Campbell. A demonstration of the visual importance and flexibility of spatial-frequency amplitude and phase.Perception, 11(3):337–346, 1982

    1982

  22. [30]

    Connect later: improving fine-tuning for robustness with targeted augmentations

    Helen Qu and Sang Michael Xie. Connect later: improving fine-tuning for robustness with targeted augmentations. InProceedings of the 41st International Conference on Machine Learning, pages 41769–41786, 2024

    2024

  23. [31]

    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, et al. Learning transferable visual models from natural language supervision. InInternational conference on machine learning, pag...

    2021

  24. [32]

    Data augmentation can improve robustness.Advances in neural information processing systems, 34:29935–29948, 2021

    Sylvestre-Alvise Rebuffi, Sven Gowal, Dan Andrei Calian, Florian Stimberg, Olivia Wiles, and Timothy A Mann. Data augmentation can improve robustness.Advances in neural information processing systems, 34:29935–29948, 2021

    2021

  25. [33]

    Extending the wilds bench- mark for unsupervised adaptation

    Shiori Sagawa, Pang Wei Koh, Tony Lee, Irena Gao, Sang Michael Xie, Kendrick Shen, Ananya Kumar, Weihua Hu, Michihiro Yasunaga, Henrik Marklund, et al. Extending the wilds bench- mark for unsupervised adaptation. InInternational Conference on Learning Representations, 2022. 11

    2022

  26. [34]

    Noisereduce: Domain general noise reduction for time series signals.arXiv preprint arXiv:2412.17851, 2024

    Tim Sainburg and Asaf Zorea. Noisereduce: Domain general noise reduction for time series signals.arXiv preprint arXiv:2412.17851, 2024

    2024 arXiv

  27. [35]

    Demystifying galaxy classification: An elegant and powerful hybrid approach

    Ankita Sarkar, Sarbani Palit, and Ujjwal Bhattacharya. Demystifying galaxy classification: An elegant and powerful hybrid approach. In2024 39th International Conference on Image and Vision Computing New Zealand (IVCNZ), pages 1–6. IEEE, 2024

    2024

  28. [36]

    Connect, not collapse: Explaining contrastive learning for unsupervised domain adaptation

    Kendrick Shen, Robbie M Jones, Ananya Kumar, Sang Michael Xie, Jeff Z HaoChen, Tengyu Ma, and Percy Liang. Connect, not collapse: Explaining contrastive learning for unsupervised domain adaptation. InInternational conference on machine learning, pages 19847–19878. PMLR, 2022

    2022

  29. [37]

    Best practices for convolutional neural networks applied to visual document analysis

    Patrice Y Simard, David Steinkraus, John C Platt, et al. Best practices for convolutional neural networks applied to visual document analysis. InIcdar, volume 3. Edinburgh, 2003

    2003

  30. [38]

    Fixmatch: Simplifying semi- supervised learning with consistency and confidence.Advances in neural information processing systems, 33:596–608, 2020

    Kihyuk Sohn, David Berthelot, Nicholas Carlini, Zizhao Zhang, Han Zhang, Colin A Raf- fel, Ekin Dogus Cubuk, Alexey Kurakin, and Chun-Liang Li. Fixmatch: Simplifying semi- supervised learning with consistency and confidence.Advances in neural information processing systems, 33...

    2020

  31. [39]

    Correlation alignment for unsupervised domain adaptation.Domain adaptation in computer vision applications, pages 153–171, 2017

    Baochen Sun, Jiashi Feng, and Kate Saenko. Correlation alignment for unsupervised domain adaptation.Domain adaptation in computer vision applications, pages 153–171, 2017

    2017

  32. [40]

    Deep coral: Correlation alignment for deep domain adaptation

    Baochen Sun and Kate Saenko. Deep coral: Correlation alignment for deep domain adaptation. InComputer vision–ECCV 2016 workshops: Amsterdam, the Netherlands, October 8-10 and 15-16, 2016, proceedings, part III 14, pages 443–450. Springer, 2016

    2016

  33. [41]

    David Tellez, Maschenka Balkenhol, Irene Otte-Höller, Rob Van De Loo, Rob V ogels, Peter Bult, Carla Wauters, Willem Vreuls, Suzanne Mol, Nico Karssemeijer, et al. Whole-slide mitosis detection in h&e breast histology using phh3 as a reference to train distilled stain-invarian...

    2018

  34. [42]

    Galaxy morphology classification with densenet

    Guangze Wang. Galaxy morphology classification with densenet. InJournal of Physics: Conference Series, volume 2580, page 012064. IOP Publishing, 2023

    2023

  35. [43]

    A fine-grained analysis on distribution shift

    Olivia Wiles, Sven Gowal, Florian Stimberg, Sylvestre-Alvise Rebuffi, Ira Ktena, Krishna- murthy Dj Dvijotham, and Ali Taylan Cemgil. A fine-grained analysis on distribution shift. In International Conference on Learning Representations, 2021

    2021

  36. [44]

    Unsupervised data augmenta- tion for consistency training.Advances in neural information processing systems, 33:6256–6268, 2020

    Qizhe Xie, Zihang Dai, Eduard Hovy, Thang Luong, and Quoc Le. Unsupervised data augmenta- tion for consistency training.Advances in neural information processing systems, 33:6256–6268, 2020

    2020

  37. [45]

    Self-training with noisy student improves imagenet classification

    Qizhe Xie, Minh-Thang Luong, Eduard Hovy, and Quoc V Le. Self-training with noisy student improves imagenet classification. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 10687–10698, 2020

    2020

  38. [46]

    Fourier-based augmentation with applications to domain generalization.Pattern Recognition, 139:109474, 2023

    Qinwei Xu, Ruipeng Zhang, Ziqing Fan, Yanfeng Wang, Yi-Yan Wu, and Ya Zhang. Fourier-based augmentation with applications to domain generalization.Pattern Recognition, 139:109474, 2023

    2023

  39. [47]

    Improve unsupervised domain adaptation with mixup training.arXiv preprint arXiv:2001.00677, 2020

    Shen Yan, Huan Song, Nanxiang Li, Lincan Zou, and Liu Ren. Improve unsupervised domain adaptation with mixup training.arXiv preprint arXiv:2001.00677, 2020

    2001 arXiv

  40. [48]

    Interactive self-training with mean teachers for semi-supervised object detection

    Qize Yang, Xihan Wei, Biao Wang, Xian-Sheng Hua, and Lei Zhang. Interactive self-training with mean teachers for semi-supervised object detection. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 5941–5950, 2021

    2021

  41. [49]

    Phase consistent ecological domain adaptation

    Yanchao Yang, Dong Lao, Ganesh Sundaramoorthi, and Stefano Soatto. Phase consistent ecological domain adaptation. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 9011–9020, 2020

    2020

  42. [50]

    Fda: Fourier domain adaptation for semantic segmentation

    Yanchao Yang and Stefano Soatto. Fda: Fourier domain adaptation for semantic segmentation. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 4085–4095, 2020. 12

    2020

  43. [51]

    Im- proving out-of-distribution robustness via selective augmentation

    Huaxiu Yao, Yu Wang, Sai Li, Linjun Zhang, Weixin Liang, James Zou, and Chelsea Finn. Im- proving out-of-distribution robustness via selective augmentation. InInternational Conference on Machine Learning, pages 25407–25437. PMLR, 2022

    2022

  44. [52]

    Cutmix: Regularization strategy to train strong classifiers with localizable features

    Sangdoo Yun, Dongyoon Han, Seong Joon Oh, Sanghyuk Chun, Junsuk Choe, and Youngjoon Yoo. Cutmix: Regularization strategy to train strong classifiers with localizable features. In Proceedings of the IEEE/CVF international conference on computer vision, pages 6023–6032, 2019

    2019

  45. [53]

    mixup: Beyond empirical risk minimization.arXiv preprint arXiv:1710.09412, 2017

    Hongyi Zhang, Moustapha Cisse, Yann N Dauphin, and David Lopez-Paz. mixup: Beyond empirical risk minimization.arXiv preprint arXiv:1710.09412, 2017

    2017 arXiv

  46. [54]

    Deep domain-adversarial image generation for domain generalisation

    Kaiyang Zhou, Yongxin Yang, Timothy Hospedales, and Tao Xiang. Deep domain-adversarial image generation for domain generalisation. InProceedings of the AAAI conference on artificial intelligence, volume 34, pages 13025–13032, 2020. 13 A Related Work A.1 Pretraining for OOD Rob...

    2020

  47. [55]

    This underscores the importance of pretrained representations in achieving robustness gains

    show that a self-training baseline using pseudo-labels generated from an ERM model with targeted augmentations still underperforms compared to Connect Later [30]. This underscores the importance of pretrained representations in achieving robustness gains. Connect Later [30] le...

  48. [56]

    Xu et al

    introduce phase consistency as a constraint during source-target translation, demonstrating its superiority over traditional cycle consistency [17] in domain adaptation tasks, especially for semantic segmentation. Xu et al. [46] incorporates a Fourier-based strategy that empha...

Pith tools

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