Pith. sign in

REVIEW 4 major objections 7 minor 50 references

DDFP: Data-dependent Frequency Prompt for Source Free Domain Adaptation of Medical Image Segmentation

T0 review · 4 major / 7 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 This paper claims that a source-free domain adaptation pipeline combining batch-normalization preadaptation, data-dependent frequency prompts, and style-layer-only fine-tuning outperforms existing methods on cross-modality medical image…

desk verdict A genuinely new per-image frequency prompt for SFDA, but the headline SOTA comparisons are not yet controlled: FVP/FSM baselines come from another split and no error bars are reported. read the letter →

arxiv 2505.09927 v1 pith:BPB3ZRHW submitted 2025-05-15 cs.CV

Siqi Yin , Shaolei Liu , Manning Wang This is my paper · ORCID
classification cs.CV
keywords source-freedomainadaptationmedicalimagesegmentationfrequencypromptdata-dependentbatchnormalizationcalibrationpseudo-labelingcross-modalitystyle-relatedlayerfine-tuning
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 tries to establish that source-free domain adaptation (SFDA) for medical image segmentation works best when the domain gap is attacked at three different moments: recalibrate the source model's batch-normalization statistics on target data before training begins; translate each target image into a source-like style with a per-image, data-dependent prompt in the frequency domain; and fine-tune only the shallow style-related layers of the target network. The claimed payoff is that this combination outperforms existing SFDA methods on cross-modality abdominal, cardiac, and brain tumor segmentation, including an average Dice of 0.8987 versus 0.8382 for the best prior method on abdominal CT-to-MRI with a U-Net. A curious reader would care because in medical settings the labeled source data may be unavailable for privacy reasons, and the paper argues the previous fixes—one fixed style prompt for the whole domain and whole-model fine-tuning—leave both style transfer and training efficiency on the table.

What carries the argument

The load-bearing mechanism is the data-dependent frequency prompt generation module. For each target image it takes the FFT amplitude and phase spectra, processes them through separate 1x1-convolution networks, concatenates the features with a learnable domain-dependent frequency prompt, passes them through a fusion network, and extracts a per-image prompt that is blended with the shared prompt via a skip connection. This prompt is applied multiplicatively to the amplitude spectrum, and an inverse FFT reconstructs a source-like target image. The other two carrying components are the BN preadaptation step, which recalibrates the source model's running statistics on target data before any training, and the style-related layer fine-tuning strategy, which freezes everything except the first four convolutional layers and the prompt parameters.

What would settle it

Compare DDFP against an adaptation method that trains the full network using the same pseudo-labels and prompted images on a dataset whose domain shift includes anatomical shape or label-distribution changes; if full-network training clearly wins, the fixed shallow/deep split is the limiting assumption. Concretely, on a target dataset with substantially different organ proportions or pathology, compute Dice of the frozen-deep version against the same method with deeper layers trainable.

Watch

Extended reading notes

Core claim

The paper's central claim is that a single domain-level style prompt is the bottleneck, and that conditioning the prompt on each image's Fourier amplitude and phase spectra removes that bottleneck. The DDFP module computes a shared trainable frequency prompt and then adjusts it per image through small 1x1-convolution networks before multiplying the result onto the image's amplitude spectrum and reconstructing the image with an inverse FFT. The same framework gets a second boost by using a batch-normalization-preadapted model, rather than the raw source model, to initialize the target model and to produce filtered pseudo-labels. With only the first four convolutional layers trainable, the paper reports top average Dice on both abdominal and cardiac cross-modality tasks and superior results on brain tumor Flair/T2 adaptation.

Load-bearing premise

The whole pipeline assumes the domain gap lives in low-level style and that shallow layers (the first four convolutions of a U-Net) carry it while deep content layers can be frozen, so if a shift also alters content-level features the frozen layers cannot be repaired.

Editorial extensions

If this is right

  • On the paper's own numbers, a fixed source model can reach average Dice around 0.90 on abdominal CT-to-MRI and 0.85 on cardiac MRI-to-CT without ever seeing source data again, provided target-model training uses prompted images and reliable pseudo-labels.
  • Adaptation cost drops because only a few shallow layers and small prompt modules are trainable, so the trainable FLOPs are much lower than whole-model fine-tuning.
  • The gain should be largest when the domain gap is large: BN preadaptation gives a big boost in hard directions such as abdominal MRI-to-CT and a small boost in easy directions.
  • A single shared domain prompt is not enough; the paper's ablation suggests conditioning the prompt on each image's amplitude and phase spectra is what closes the remaining gap.
  • The style/content layer split means the method should work best for shifts that are mostly low-level appearance changes, which is precisely the cross-modality setting tested here.

Reading between the lines

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

  • The paper fixes the first four convolutional layers as style-related; the natural follow-up I would draw is to make that split adaptive, for example by selecting trainable layers from target-data BN-statistic drift or gradient sensitivity, so the method could handle shifts that penetrate deeper into the network.
  • Because the prompts act on the low-frequency part of the amplitude spectrum, a minimal testable baseline would be a per-image histogram or affine intensity matching in the frequency domain; if that baseline matched DDFP, the learned prompt's advantage would narrow considerably.
  • The paper evaluates Dice and ASD and notes blurred boundaries, so an implicit extension is to add boundary- or contour-aware supervision to the prompted-image pipeline, which could improve the ASD numbers without changing the core mechanism.
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 / 7 minor

Summary. The manuscript proposes DDFP, a source-free domain adaptation (SFDA) framework for medical image segmentation. The method combines three components: a momentum-based batch-normalization preadaptation of the source model using target data; a data-dependent frequency prompt generated by a small network that modulates the amplitude spectrum of each input image; and fine-tuning of only the first four convolutional layers of the U-Net backbone while freezing deeper layers. The target model is trained with a weighted combination of pseudo-label, BN-statistic, and entropy losses. Experiments on abdominal (CHAOS/Multi-Atlas), cardiac (MMWHS), and brain tumor (BraTS) datasets report Dice and ASD improvements over several SFDA baselines, with the largest gains on cross-modality abdominal CT-to-MRI. Ablations examine loss components, prompt design, BN preadaptation, and the number of trainable layers.

Significance. If the controlled comparisons were solid, the paper would be a useful contribution: the per-image data-dependent frequency prompt is a genuine departure from domain-level prompts in FVP/FSM, and the BN preadaptation step is a lightweight way to improve pseudo-label quality. The ablation studies are informative, and Fig. 9 provides a partial validation of the style-layer fine-tuning strategy. However, the headline claim of outperforming state-of-the-art methods is currently supported only by uncontrolled baseline numbers for FVP and FSM, and by single-run evaluations without error bars. The empirical significance is therefore conditional on the authors redoing the comparisons under a common protocol.

major comments (4)
  1. [4.1.4, Tables 1-2] The results for FVP and FSM are taken from [38] (indicated by *), which used a random data partition, whereas all DDFP numbers come from a different 8:2 split of the same datasets. Consequently, per-table comparisons such as the DeepLab CT-to-MRI row (DDFP 0.8501 vs. FVP 0.7330) are not controlled experiments: differences could be due to which slices were used for testing rather than to the method. Please re-run these baselines under the identical split, or remove the starred rows from the headline SOTA claims and qualify the comparisons accordingly.
  2. [3.5, Eq. (12)] The pseudo-label loss is written with a confidence weight conf^{G_t}_{h,w}, while the preceding text and the pseudo-label selection in Eq. (9) use the preadapted model G'_s. Using the target model's own confidence for reweighting would be a different training signal, and the notation is inconsistent. Please correct the equation and state explicitly which model produces the confidence weights.
  3. [3.6, 4.1.3] The trainable-layer selection is specified as the first four convolutional layers for the U-Net backbone, but the paper also reports DeepLabv3 results. The analogous style-related layers for ResNet50/DeepLabv3 are never defined, so the DeepLab rows in Tables 1-3 cannot be reproduced. Please specify the layer split for each backbone. I do not regard the heuristic nature of the style/content split itself as disqualifying, since Fig. 9 explores alternate layer selections, but the missing DeepLab specification is a concrete reproducibility gap.
  4. [4.1.3, 4.5] All quantitative results appear to be single-run; no error bars, seeds, or per-comparison statistical tests are given. The Wilcoxon test in Fig. 6 aggregates 3D Dice over four tasks with roughly 5 test volumes per task, which is not a substitute for per-table significance tests under matched test slices. Please provide at least three seeds with standard deviations, and clarify whether the aggregated test uses paired samples from the same test split.
minor comments (7)
  1. [Abstract, Section 1] The abstract and introduction state that code is available online, but no URL is provided; please include a repository link.
  2. [4.6.2, Table 5] The text says that two sets of experiments (with and without fine-tuning of style-related layers) were conducted, but Table 5 reports only one set; the additional results should be shown or the sentence corrected.
  3. [4.4] The method name is written 'TT-FSUDA' in the brain tumor section, while the rest of the paper uses 'TT-SFUDA'; please unify the naming.
  4. [4.6.2, 4.6.3] Both Section 4.6.2 and Section 4.6.3 refer to 'Fig. 8' for different content; renumber the figures and update the cross-references.
  5. [4.6.3] The phrase 'pseudo-label generalization' should read 'pseudo-label generation.'
  6. [3.4] The module name is written inconsistently as both 'DDFP' and 'DFFP'; please adopt one consistent abbreviation.
  7. [4.5] In Section 4.5, 'restuls' is a typo for 'results.'

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the central claim is an empirical comparison on held-out public benchmark labels, and all self-referential training signals are adaptation objectives, not evaluation outcomes.

full rationale

The paper's claimed contribution is an SFDA framework whose success is measured by Dice and ASD on held-out ground-truth labels from public benchmarks (CHAOS, MMWHS, BraTS2018) using an 8:2 split. The only quantities derived from the source model and unlabeled target images are training signals: BN pre-adaptation (Eq. 1) recalibrates source BN statistics with target statistics, the frequency prompt is optimized with the BN-statistic loss (Eq. 10), entropy minimization (Eq. 11), and pseudo-label loss (Eq. 12), and pseudo-labels come from the preadapted source model. These are self-supervised objectives that define the adaptation procedure; they do not by construction determine the reported test-set Dice or ASD numbers, which come from external manual labels. The style/content layer split is an explicitly stated design choice ('we designate the first four convolutional layers of our U-Net backbone as the style-related layers, which are trainable'), not a fitted parameter or a result smuggled in by citation. The paper cites prior work by overlapping authors ([23], [24]) only for preprocessing conventions and context, not as the load-bearing justification for the main result. The most serious validity concern is comparison fairness—FSM* and FVP* results are quoted from [38], which used a different random partition—but that is a benchmarking/reproducibility risk, not circularity. No step in the derivation reduces to its own inputs by construction, so the circularity score is 0.

Assumptions & free parameters 8 free parameters · 4 assumptions · 1 invented entities

The central claim rests on standard FFT operations and on two domain assumptions (amplitude=style, shallow layers=style) inherited from prior work, plus a set of hand-tuned hyperparameters. No new physical entities are posited; the only invented module is the learned prompt generator, whose benefit is demonstrated only internally.

free parameters (8)
  • Fusion weight alpha = 0.2
    Controls skip connection between domain-dependent and data-dependent prompts in Eq. 4; chosen by hand (Section 4.1.3), ablated with a claim of minimal impact.
  • Pseudo-label class retention ratio delta_cls = 40%
    Fraction of lowest-entropy pixels retained per class (Eq. 7); set to 40 for all classes (Section 4.1.3). Text later inconsistently says 0.4.
  • Global entropy threshold delta_glo = 0.4
    Global entropy cutoff for pseudo-label reliability (Eq. 8); set to 0.4.
  • BN preadaptation momentum rho = 0.1
    Momentum in Eq. 1 for updating source BN statistics with target batches.
  • BN preadaptation epochs E_W = 10
    Number of forward-propagation epochs over target data for BN recalibration.
  • Pseudo-label loss scale upsilon = 0.2
    Hyperparameter scaling the pseudo-label loss in Eq. 12.
  • Per-task loss weights [w_ent, w_BNS, w_pseu] = e.g., [1,1,10], [0.1,1,10], [4,0.1,10], etc.
    Hand-tuned per dataset and adaptation direction (Section 4.1.3) to normalize loss magnitudes; a form of per-task fitting.
  • Number of style-related layers = first 4 conv layers of U-Net
    Design choice for which layers are trainable; based on style/content layer assumption, ablated in Fig. 9.
assumptions (4)
  • domain assumption Amplitude spectrum of Fourier transform encodes style/grayscale, phase encodes structure/content
    Used throughout Section 3.4; prompt is applied only to amplitude, leaving phase unchanged, motivated by prior FDA/FVP results.
  • domain assumption Shallow network layers encode style and deep layers encode content, so fine-tuning only the first four conv layers is sufficient for domain adaptation
    Invoked in Section 3.6 and Problem 3 in the Introduction; cites [30], [6].
  • domain assumption BN statistic matching between source and target models is a valid style alignment objective
    Used in the BNS loss Eq. 10; relies on BN layers capturing domain-specific statistics, a standard assumption in TTA/SFDA literature.
  • standard math FFT/inverse FFT and 1x1 conv modules operate as described without numerical instability
    Standard signal processing and network operations; no special conditions needed.
invented entities (1)
  • Data-dependent frequency prompt (DDFP) generation module G_DDFP
    purpose: Produces a per-image frequency-domain multiplier applied to the amplitude spectrum to translate target images into source-like style.
    Evidence comes only from internal ablation (Table 5); there is no external, independently verifiable prediction made by the entity.

how reviews work

0 comments
Cite this review

Pith. "Pith review of DDFP: Data-dependent Frequency Prompt for Source Free Domain Adaptation of Medical Image Segmentation." pith.science (2026). https://pith.science/paper/BPB3ZRHW

@misc{pith2026250509927,
  author       = {Pith},
  title        = {Pith review of: DDFP: Data-dependent Frequency Prompt for Source Free Domain Adaptation of Medical Image Segmentation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/BPB3ZRHW}},
  note         = {Machine review of arXiv:2505.09927}
}
read the original abstract

Domain adaptation addresses the challenge of model performance degradation caused by domain gaps. In the typical setup for unsupervised domain adaptation, labeled data from a source domain and unlabeled data from a target domain are used to train a target model. However, access to labeled source domain data, particularly in medical datasets, can be restricted due to privacy policies. As a result, research has increasingly shifted to source-free domain adaptation (SFDA), which requires only a pretrained model from the source domain and unlabeled data from the target domain data for adaptation. Existing SFDA methods often rely on domain-specific image style translation and self-supervision techniques to bridge the domain gap and train the target domain model. However, the quality of domain-specific style-translated images and pseudo-labels produced by these methods still leaves room for improvement. Moreover, training the entire model during adaptation can be inefficient under limited supervision. In this paper, we propose a novel SFDA framework to address these challenges. Specifically, to effectively mitigate the impact of domain gap in the initial training phase, we introduce preadaptation to generate a preadapted model, which serves as an initialization of target model and allows for the generation of high-quality enhanced pseudo-labels without introducing extra parameters. Additionally, we propose a data-dependent frequency prompt to more effectively translate target domain images into a source-like style. To further enhance adaptation, we employ a style-related layer fine-tuning strategy, specifically designed for SFDA, to train the target model using the prompted target domain images and pseudo-labels. Extensive experiments on cross-modality abdominal and cardiac SFDA segmentation tasks demonstrate that our proposed method outperforms existing state-of-the-art methods.

Figures

Figures reproduced from arXiv: 2505.09927 by the authors.

Figure 1
Figure 1. Comparison between (a) the previous source-free domain adaptation framework and (b) the proposed DDFP framework. The DDFP framework reduces the domain gap throughout both the initial and subsequent training phases by utilizing a preadapted model, data-dependent frequency prompt learning, and pseudo-labeling strategies. S. Yin, S. Liu, M. Wang: Preprint submitted to Elsevier Page 13 of 13 [PITH_FULL_IMAGE:figures/fu… view at source ↗
Figure 2
Figure 2. Overview of the proposed DDFP architecture. We introduce a BN preadaptation strategy (yellow) to initialize the target model and generate high-quality pseudo-labels for target domain data (blue). The data-dependent frequency prompt learning strategy (red) facilitates image style translation. Both the data-dependent frequency prompt parameters and the style-related layers of the target model are jointly trained to ac… view at source ↗
Figure 3
Figure 3. Data-dependent frequency prompt generation process for each image in the target domain batch. S. Yin, S. Liu, M. Wang: Preprint submitted to Elsevier Page 14 of 13 [PITH_FULL_IMAGE:figures/full_fig_p014_3.png] view at source ↗
Figures from the paper (4 more)
Figure 4
Figure 4. Figure 4: Visualization of SFDA segmentation results on the multiorgan abdominal dataset. The first two rows show the results for CT to MRI adaptation, while the last two rows display results for MRI to CT adaptation. Image W/o adaptation CBMT ProContra TT-SFUDA DDFP (ours) Supe…
Figure 5
Figure 5. Figure 5: Visualization of SFDA segmentation results on the cardiac dataset. The first two rows correspond to CT to MRI adaptation, while the last two rows correspond to MRI to CT adaptation. S. Yin, S. Liu, M. Wang: Preprint submitted to Elsevier Page 15 of 13 [PITH_FULL_IMAGE…
Figure 6
Figure 6. Figure 6: The boxplot results of experiments on the abdominal and cardiac datsets. *: 1.00e-02 < p <= 5.00e-02, **: 1.00e-03 < p <= 1.00e-02, **: 1.00e-04 < p <= 1.00e-03 [PITH_FULL_IMAGE:figures/full_fig_p016_6.png]
Figure 7
Figure 7. Figure 7: Results of different target model initialization ap￾proaches. Pink: Using the BN preadapted model for target model initialization and pseudo-label generation. Blue: Using the source model for target model initialization and pseudo￾label generation. Gray: W/o adaptation…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

50 extracted references · 37 canonical work pages

  1. [38]

    Fvp:Fouriervisualpromptingforsource-freeunsupervised domain adaptation of medical image segmentation

    Wang,Y.,Cheng,J.,Chen,Y.,Shao,S.,Zhu,L.,Wu,Z.,Liu,T.,Zhu, H.,2023. Fvp:Fouriervisualpromptingforsource-freeunsupervised domain adaptation of medical image segmentation. IEEE Transac- tions on Medical Imaging

    work page 2023

  2. [1]

    Source-free domain adaptation for image segmentation

    Bateson, M., Kervadec, H., Dolz, J., Lombaert, H., Ayed, I.B., 2022. Source-free domain adaptation for image segmentation. Medical Image Analysis 82, 102617

    work page 2022

  3. [2]

    Chen, C., Dou, Q., Chen, H., Qin, J., Heng, P.A., 2019. Synergistic image and feature adaptation: Towards cross-modality domain adap- tation for medical image segmentation, in: Proceedings of the AAAI conference on artificial intelligence, pp. 865–872

    work page 2019

  4. [3]

    Unsupervised bidirectional cross-modality adaptation via deeply synergistic image and feature alignment for medical image segmentation

    Chen,C.,Dou,Q.,Chen,H.,Qin,J.,Heng,P.A.,2020. Unsupervised bidirectional cross-modality adaptation via deeply synergistic image and feature alignment for medical image segmentation. IEEE trans- actions on medical imaging 39, 2494–2505

    work page 2020

  5. [4]

    Chen, C., Liu, Q., Jin, Y., Dou, Q., Heng, P.A., 2021. Source- free domain adaptive fundus image segmentation with denoised pseudo-labeling, in: Medical Image Computing and Computer As- sisted Intervention–MICCAI 2021: 24th International Conference, Strasbourg,France,September27–October1,2021,Proceedings,Part V 24, Springer. pp. 225–235

    work page 2021

  6. [5]

    Domain adaptation in computer vision applications

    Csurka, G., et al., 2017. Domain adaptation in computer vision applications. volume 2. Springer

    work page 2017

  7. [6]

    Pnp-adanet: Plug-and-play adversarial domain adaptation network at unpaired cross-modality cardiac segmentation

    Dou, Q., Ouyang, C., Chen, C., Chen, H., Glocker, B., Zhuang, X., Heng, P.A., 2019. Pnp-adanet: Plug-and-play adversarial domain adaptation network at unpaired cross-modality cardiac segmentation. IEEE Access 7, 99065–99076

    work page 2019

  8. [7]

    Visual prompt tuning for test-time domain adaptation

    Gao, Y., Shi, X., Zhu, Y., Wang, H., Tang, Z., Zhou, X., Li, M., Metaxas, D.N., 2022. Visual prompt tuning for test-time domain adaptation. arXiv preprint arXiv:2210.04831

    arXiv 2022

Show all 50 references
  1. [8]

    Progressive transferlearningandadversarialdomainadaptationforcross-domain skin disease classification

    Gu, Y., Ge, Z., Bonnington, C.P., Zhou, J., 2020. Progressive transferlearningandadversarialdomainadaptationforcross-domain skin disease classification. IEEE Journal of Biomedical and Health Informatics 24, 1379–1393. doi:10.1109/JBHI.2019.2942429

    2020

  2. [9]

    Mcdcd: Multi-source unsuper- vised domain adaptation for abnormal human gait detection

    Guo, Y., Gu, X., Yang, G.Z., 2021. Mcdcd: Multi-source unsuper- vised domain adaptation for abnormal human gait detection. IEEE JournalofBiomedicalandHealthInformatics25,4017–4028. doi:10. 1109/JBHI.2021.3080502

    2021

  3. [10]

    Deep residual learning for image recognition, in: Proceedings of the IEEE conference on computer vision and pattern recognition, pp

    He, K., Zhang, X., Ren, S., Sun, J., 2016. Deep residual learning for image recognition, in: Proceedings of the IEEE conference on computer vision and pattern recognition, pp. 770–778

    2016

  4. [11]

    Source-free unsupervised domain adaptation for cross-modality abdominal multi-organ seg- mentation

    Hong, J., Zhang, Y.D., Chen, W., 2022. Source-free unsupervised domain adaptation for cross-modality abdominal multi-organ seg- mentation. Knowledge-Based Systems 250, 109155. S. Yin, S. Liu, M. Wang:Preprint submitted to ElsevierPage 11 of 13 DDFP: Data-dependent Frequency Pr...

    2022

  5. [12]

    Hsu, J., Chiu, W., Yeung, S., 2021. Darcnn: Domain adaptive region-based convolutional neural network for unsupervised instance segmentationinbiomedicalimages,in:ProceedingsoftheIEEE/CVF conference on computer vision and pattern recognition, pp. 1003– 1012

    2021

  6. [13]

    Prosfda: Prompt learning based source-freedomainadaptationformedicalimagesegmentation.arXiv preprint arXiv:2211.11514

    Hu, S., Liao, Z., Xia, Y., 2022. Prosfda: Prompt learning based source-freedomainadaptationformedicalimagesegmentation.arXiv preprint arXiv:2211.11514

    2022 arXiv

  7. [14]

    Fsdr: Frequency space domain randomization for domain generalization, in: Proceedings of theIEEE/CVFConferenceonComputerVisionandPatternRecogni- tion, pp

    Huang, J., Guan, D., Xiao, A., Lu, S., 2021. Fsdr: Frequency space domain randomization for domain generalization, in: Proceedings of theIEEE/CVFConferenceonComputerVisionandPatternRecogni- tion, pp. 6891–6902

    2021

  8. [15]

    Epithelium-stroma classification via convolutional neural networks and unsupervised domain adaptation in histopathological images

    Huang, Y., Zheng, H., Liu, C., Ding, X., Rohde, G.K., 2017. Epithelium-stroma classification via convolutional neural networks and unsupervised domain adaptation in histopathological images. IEEE Journal of Biomedical and Health Informatics 21, 1625–1632. doi:10.1109/JBHI.2017.2691738

    2017

  9. [16]

    Visual prompt tuning, in: European Conference on Computer Vision, Springer

    Jia,M.,Tang,L.,Chen,B.C.,Cardie,C.,Belongie,S.,Hariharan,B., Lim, S.N., 2022. Visual prompt tuning, in: European Conference on Computer Vision, Springer. pp. 709–727

    2022

  10. [17]

    Chaos challenge-combined (ct-mr) healthy abdominal organ segmentation

    Kavur,A.E.,Gezer,N.S.,Barış,M.,Aslan,S.,Conze,P.H.,Groza,V., Pham, D.D., Chatterjee, S., Ernst, P., Özkan, S., et al., 2021. Chaos challenge-combined (ct-mr) healthy abdominal organ segmentation. Medical Image Analysis 69, 101950

    2021

  11. [18]

    Miccai multi-atlas labeling beyond the cranial vault– workshop and challenge, in: Proc

    Landman, B., Xu, Z., Igelsias, J., Styner, M., Langerak, T., Klein, A., 2015. Miccai multi-atlas labeling beyond the cranial vault– workshop and challenge, in: Proc. MICCAI Multi-Atlas Labeling Beyond Cranial Vault—Workshop Challenge, p. 12

    2015

  12. [19]

    Dataset shift in machine learning

    LEARNING, T.S.I.M., . Dataset shift in machine learning

  13. [20]

    A comprehensive survey on source-free domain adaptation

    Li, J., Yu, Z., Du, Z., Zhu, L., Shen, H.T., 2024. A comprehensive survey on source-free domain adaptation. IEEE Transactions on Pattern Analysis and Machine Intelligence

    2024

  14. [21]

    9641–9650

    Li,R.,Jiao,Q.,Cao,W.,Wong,H.S.,Wu,S.,2020.Modeladaptation: Unsuperviseddomainadaptationwithoutsourcedata,in:Proceedings of the IEEE/CVF conference on computer vision and pattern recog- nition, pp. 9641–9650

    2020

  15. [22]

    Revisiting batch normalization for practical domain adaptation

    Li, Y., Wang, N., Shi, J., Liu, J., Hou, X., 2016. Revisiting batch normalization for practical domain adaptation. arXiv preprint arXiv:1603.04779

    2016 arXiv

  16. [23]

    Liu, S., Yin, S., Qu, L., Wang, M., 2023a. Reducing domain gap in frequency and spatial domain for cross-modality domain adaptation on medical image segmentation, in: Proceedings of the AAAI Con- ference on Artificial Intelligence, pp. 1719–1727

  17. [24]

    A structure- aware framework of unsupervised cross-modality domain adaptation via frequency and spatial knowledge distillation

    Liu, S., Yin, S., Qu, L., Wang, M., Song, Z., 2023b. A structure- aware framework of unsupervised cross-modality domain adaptation via frequency and spatial knowledge distillation. IEEE Transactions on Medical Imaging

  18. [25]

    Source-freedomainadaptationfor semanticsegmentation,in:ProceedingsoftheIEEE/CVFConference on Computer Vision and Pattern Recognition, pp

    Liu,Y.,Zhang,W.,Wang,J.,2021. Source-freedomainadaptationfor semanticsegmentation,in:ProceedingsoftheIEEE/CVFConference on Computer Vision and Pattern Recognition, pp. 1215–1224

    2021

  19. [26]

    Margin preservingself-pacedcontrastivelearningtowardsdomainadaptation for medical image segmentation

    Liu, Z., Zhu, Z., Zheng, S., Liu, Y., Zhou, J., Zhao, Y., 2022. Margin preservingself-pacedcontrastivelearningtowardsdomainadaptation for medical image segmentation. IEEE Journal of Biomedical and Health Informatics 26, 638–647. doi:10.1109/JBHI.2022.3140853

    2022

  20. [27]

    The multimodal brain tumor image segmentation benchmark (brats)

    Menze,B.H.,Jakab,A.,Bauer,S.,Kalpathy-Cramer,J.,Farahani,K., Kirby, J., Burren, Y., Porz, N., Slotboom, J., Wiest, R., et al., 2014. The multimodal brain tumor image segmentation benchmark (brats). IEEE transactions on medical imaging 34, 1993–2024

    2014

  21. [28]

    Source-free online domain adaptive semantic segmentation of satellite images under image degradation

    Niloy, F.F., Bhaumik, K.K., Woo, S.S., 2024. Source-free online domain adaptive semantic segmentation of satellite images under image degradation. arXiv preprint arXiv:2401.02113

    2024 arXiv

  22. [29]

    Blackvip: Black-box visual prompting for robust transfer learning, in: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp

    Oh, C., Hwang, H., Lee, H.y., Lim, Y., Jung, G., Jung, J., Choi, H., Song, K., 2023. Blackvip: Black-box visual prompting for robust transfer learning, in: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 24224–24235

    2023

  23. [30]

    Two at once: Enhancing learning and generalization capacities via ibn-net, in: Proceedings of the european conference on computer vision (ECCV), pp

    Pan, X., Luo, P., Shi, J., Tang, X., 2018. Two at once: Enhancing learning and generalization capacities via ibn-net, in: Proceedings of the european conference on computer vision (ECCV), pp. 464–479

    2018

  24. [31]

    Curriculum feature alignment domain adaptation for epithelium-stroma classification in histopathological images

    Qi, Q., Lin, X., Chen, C., Xie, W., Huang, Y., Ding, X., Liu, X., Yu, Y., 2021. Curriculum feature alignment domain adaptation for epithelium-stroma classification in histopathological images. IEEE JournalofBiomedicalandHealthInformatics25,1163–1172. doi:10. 1109/JBHI.2020.3021558

    2021

  25. [32]

    Ronneberger, O., Fischer, P., Brox, T., 2015. U-net: Convolutional networks for biomedical image segmentation, in: Medical image computing and computer-assisted intervention–MICCAI 2015: 18th international conference, Munich, Germany, October 5-9, 2015, pro- ceedings, part III...

    2015

  26. [33]

    Improving robustness against common corrup- tions by covariate shift adaptation

    Schneider, S., Rusak, E., Eck, L., Bringmann, O., Brendel, W., Bethge, M., 2020. Improving robustness against common corrup- tions by covariate shift adaptation. Advances in neural information processing systems 33, 11539–11551

    2020

  27. [34]

    Tang, L., Li, K., He, C., Zhang, Y., Li, X., 2023. Source-free do- main adaptive fundus image segmentation with class-balanced mean teacher, in: International Conference on Medical Image Computing and Computer-Assisted Intervention, Springer. pp. 684–694

    2023

  28. [35]

    Target and task specificsource-freedomainadaptiveimagesegmentation,in:Medical Imaging with Deep Learning

    VS, V., Valanarasu, J.M.J., Patel, V.M., 2024. Target and task specificsource-freedomainadaptiveimagesegmentation,in:Medical Imaging with Deep Learning. URL:https://openreview.net/forum? id=Ym30gCtKqN

    2024

  29. [36]

    Exploring domain- invariant parameters for source free domain adaptation, in: Proceed- ings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp

    Wang, F., Han, Z., Gong, Y., Yin, Y., 2022. Exploring domain- invariant parameters for source free domain adaptation, in: Proceed- ings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 7151–7160

    2022

  30. [37]

    Towardsgenericsemi-supervisedframework for volumetric medical image segmentation

    Wang,H.,Li,X.,2024. Towardsgenericsemi-supervisedframework for volumetric medical image segmentation. Advances in Neural Information Processing Systems 36

    2024

  31. [39]

    Calibration-freeonlinetest- timeadaptationforelectroencephalographymotorimagerydecoding, in: 2024 12th International Winter Conference on Brain-Computer Interface (BCI), IEEE

    Wimpff,M.,Döbler,M.,Yang,B.,2024. Calibration-freeonlinetest- timeadaptationforelectroencephalographymotorimagerydecoding, in: 2024 12th International Winter Conference on Brain-Computer Interface (BCI), IEEE. pp. 1–6

    2024

  32. [40]

    Test-time domainadaptationbylearningdomain-awarebatchnormalization,in: Proceedings of the AAAI Conference on Artificial Intelligence, pp

    Wu,Y.,Chi,Z.,Wang,Y.,Plataniotis,K.N.,Feng,S.,2024. Test-time domainadaptationbylearningdomain-awarebatchnormalization,in: Proceedings of the AAAI Conference on Artificial Intelligence, pp. 15961–15969

    2024

  33. [41]

    Bidirectional mapping-based domain adaptation for nucleus detection in cross- modality microscopy images

    Xing,F.,Cornish,T.C.,Bennett,T.D.,Ghosh,D.,2020. Bidirectional mapping-based domain adaptation for nucleus detection in cross- modality microscopy images. IEEE transactions on medical imaging 40, 2880–2896

    2020

  34. [42]

    Source free domain adaptationformedicalimagesegmentationwithfourierstylemining

    Yang, C., Guo, X., Chen, Z., Yuan, Y., 2022. Source free domain adaptationformedicalimagesegmentationwithfourierstylemining. Medical Image Analysis 79, 102457

    2022

  35. [43]

    Federated semi- supervised learning for covid region segmentation in chest ct using multi-national data from china, italy, japan

    Yang,D.,Xu,Z.,Li,W.,Myronenko,A.,Roth,H.R.,Harmon,S.,Xu, S., Turkbey, B., Turkbey, E., Wang, X., et al., 2021. Federated semi- supervised learning for covid region segmentation in chest ct using multi-national data from china, italy, japan. Medical image analysis 70, 101992

    2021

  36. [44]

    Yang, S., Wu, J., Liu, J., Li, X., Zhang, Q., Pan, M., Zhang, S.,

  37. [45]

    Fda: Fourier domain adaptation for semanticsegmentation,in:ProceedingsoftheIEEE/CVFconference on computer vision and pattern recognition, pp

    Yang, Y., Soatto, S., 2020. Fda: Fourier domain adaptation for semanticsegmentation,in:ProceedingsoftheIEEE/CVFconference on computer vision and pattern recognition, pp. 4085–4095

    2020

  38. [46]

    A novel 3d unsupervised domain adaptation framework for cross-modality medical image segmentation

    Yao, K., Su, Z., Huang, K., Yang, X., Sun, J., Hussain, A., Coenen, F., 2022. A novel 3d unsupervised domain adaptation framework for cross-modality medical image segmentation. IEEE Journal of BiomedicalandHealthInformatics26,4976–4986.doi:10.1109/JBHI. 2022.3162118

    2022

  39. [47]

    Source-free domain adaptation for medical image segmentation via S

    Yu, Q., Xi, N., Yuan, J., Zhou, Z., Dang, K., Ding, X., 2023. Source-free domain adaptation for medical image segmentation via S. Yin, S. Liu, M. Wang:Preprint submitted to ElsevierPage 12 of 13 DDFP: Data-dependent Frequency Prompt for Source Free Domain Adaptation of Medical...

    2023

  40. [48]

    Domainadaptor: A novel approach to test-time adaptation, in: Proceedings of the IEEE/CVF International Conference on Computer Vision, pp

    Zhang, J., Qi, L., Shi, Y., Gao, Y., 2023. Domainadaptor: A novel approach to test-time adaptation, in: Proceedings of the IEEE/CVF International Conference on Computer Vision, pp. 18971–18981

    2023

  41. [49]

    Multi-scale patch and multi-modality atlases for whole heart segmentation of mri

    Zhuang, X., Shen, J., 2016. Multi-scale patch and multi-modality atlases for whole heart segmentation of mri. Medical image analysis 31, 77–87. Source Model Target DomainSource Domain Target Model initialize Source Model Target Model Target images Pre-adapted Model Target imag...

    2016

  42. [2023]

    arXiv preprint arXiv:2303.09792

    Exploring sparse visual prompt for cross-domain semantic segmentation. arXiv preprint arXiv:2303.09792

    arXiv

Pith tools

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