Pith. sign in

REVIEW 3 major objections 5 minor 77 references

TopoTTA: Topology-Enhanced Test-Time Adaptation for Tubular Structure Segmentation

T0 review · 3 major / 5 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read TopoTTA proposes the first test-time adaptation framework tailored to tubular structure segmentation, reporting an average 31.81% clDice improvement across ten cross-domain datasets.

desk verdict First TTA method for tubular structures with broad, consistent experiments; the main open question is Stage 2 pseudo-label reliability, and the headline metric overstates the margin over the strongest baseline. read the letter →

arxiv 2508.00442 v1 pith:CCRIYO34 submitted 2025-08-01 cs.CV cs.AI

Jiale Zhou , Wenhan Wang , Shikun Li , Xiaolei Qu , Xin Guo , Yizhong Liu , Wenzhong Tang , Xun Lin
show 1 more author
Yefeng Zheng
This is my paper
classification cs.CVcs.AI
keywords test-timeadaptationtubularstructuresegmentationtopologicalcontinuityclDicecentraldifferenceconvolutionpseudo-breakgenerationdomainshiftconsistencyregularization
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

TopoTTA argues that test-time adaptation for tubular structure segmentation should optimize topology itself, not just pixel accuracy. The paper proposes a two-stage pipeline in which Stage 1 rewires the encoder's convolutions into eight directional difference operators whose combination is reweighted per test sample, and Stage 2 manufactures local pseudo-breaks in confidently segmented regions and trains the model to restore continuity against teacher pseudo-labels. Across retinal, road, microscopic-neuron, and OCT-angiography datasets, TopoTTA reports an average clDice gain of 31.81% over seven prior test-time adaptation methods, with higher Dice in most settings. If the claim holds, topology-aware self-supervision becomes a practical recipe for online adaptation of thin-structure segmenters without source data or label access.

What carries the argument

The load-bearing machinery is the pair of topology-specific modules. TopoMDC is a family of eight directional difference convolutions built by extending central difference convolution to a two-pixel difference along each of eight directions; because local tube segments are directional and elongated, the router can select the combination of directions that matches the test patch's trajectory, and resetting the router per sample lets each test image receive its own topological prior. TopoHG is the counterpart that creates targeted training signal: it turns a confident foreground patch into a pseudo-broken hard sample by low-frequency Fourier swapping with a nearby low-confidence background patch, then applies consistency regularization with a position-dependent weight map so the model learns the local foreground/background cues needed to reconnect broken predictions.

What would settle it

Take a target domain whose tubular appearance is systematically different from anything in the source training data, so the teacher produces confident but wrong pseudo-labels in whole regions; run Stage 2 alone and compare Betti errors against Stage 1 alone. If Betti errors increase or the clDice gain disappears, the pseudo-label reliability premise is what failed.

Watch

Extended reading notes

Core claim

The central claim is that a source-trained CNN can be adapted to an unseen tubular-structure image at test time by using topological structure as the adaptation signal. Stage 1 replaces each vanilla $3\times3$ convolution in the encoder with eight Topological Meta Difference Convolutions (TopoMDCs), each comparing the central pixel with one of eight neighboring directions; a learnable router reweights these operators per image patch, and only the 1,280 router parameters are updated, leaving pretrained weights untouched. Stage 2, called Topology Hard sample Generation (TopoHG), selects high-confidence foreground key points from teacher pseudo-labels, finds the neighboring background window with lowest pseudo-label confidence, swaps low-frequency Fourier components between the foreground and background patches, and overlays the edit only on pseudo-foreground pixels; the student is then aligned to the teacher's pseudo-label with a consistency loss that is up-weighted inside the pseudo-break regions, the teacher being updated by exponential moving average. The paper reports that this two-stage procedure beats seven prior test-time adaptation methods on clDice across all ten cross-domain pairings and on Dice in most settings, including cases where every comparison method degrades.

Load-bearing premise

Stage 2's continuity gain assumes the teacher's pseudo-labels are trustworthy in the high-confidence regions where pseudo-breaks are inserted; if those labels are wrong, the consistency update teaches the model to reinforce the mistake and can worsen topological continuity.

Editorial extensions

If this is right

  • Across ten cross-domain dataset pairs and two standard backbone families (plus a DSCNet variant), TopoTTA reports the highest clDice among all compared methods, with an average 31.81% improvement and gains over the second-best method of 4.66 clDice and 3.95 Dice on UNet.
  • The improvement is not bought by sacrificing segmentation accuracy: Dice also rises on most pairings, and in the DeepGlobe-to-CNDS case the method improves where every comparison method declines.
  • Because only router parameters are updated in Stage 1 and pseudo-break consistency needs no source labels in Stage 2, the pipeline fits the plug-and-play, per-sample online setting where target data arrive one image at a time.
  • The design applies to CNN-based tubular segmenters generally rather than a single architecture, which the paper demonstrates by reporting results with UNet, CS2Net, and a Stage-2-only DSCNet variant.

Reading between the lines

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

  • The paper's experiments are all in 2D; a testable extension is to lift the eight-direction TopoMDC family and the sliding-window pseudo-break search to 26-neighborhood 3D convolutions for airways or volumetric vessels, where continuity failures are arguably more damaging.
  • The contribution of TopoHG's low-frequency swap suggests that high-frequency texture carries cross-domain foreground identity; one could isolate this by ablating only the high-frequency preservation while keeping the same pseudo-break locations, rather than swapping whole augmentation schemes.
  • The teacher pseudo-labels are used only as supervision targets; the same per-patch confidence statistics that TopoHG computes (least-confident background windows near confident foreground) could be repurposed as an uncertainty or pseudo-label-quality signal for filtering, not just for hard-sample generation.
  • The router-reset design makes each test sample an independent adaptation episode, so streaming or batch settings with correlated samples may behave differently from the per-image iterations evaluated here; measuring TopoTTA on video or volume sequences would test that boundary.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 5 minor

Summary. The paper proposes TopoTTA, a two-stage test-time adaptation method for tubular structure segmentation. Stage 1 defines eight Topological Meta Difference Convolutions built from the pre-trained vanilla kernels and learns per-patch router weights (about 1,280 new parameters) via entropy minimization, leaving the pre-trained weights untouched. Stage 2 generates hard samples with local pseudo-breaks by swapping low-frequency components between foreground and nearby background windows selected from high-confidence teacher predictions, and applies teacher-student consistency (cross-entropy with a spatial weight map) to refine topological continuity. The method is evaluated on ten cross-domain source-target pairs spanning retinal vessel, road, neuronal, and OCT-A vessel segmentation, using UNet, CS2Net, and a DSCNet variant, against seven prior TTA methods. The authors report consistent gains in Dice, clDice, and Betti errors, and an average clDice improvement of 31.81%.

Significance. Test-time adaptation specifically for tubular structures is an underexplored and practically relevant problem, and the paper has several concrete strengths: the TopoMDC design is parameter-efficient, the TopoHG hard-sample generation is original, the evaluation covers more datasets and backbones than is typical in the TTA segmentation literature, the ablations isolate the two stages, and the appendix includes paired t-tests and synthesis-quality analysis. I find no circularity: the target labels come from external benchmarks, no constants are fitted to those labels, and the pseudo-label supervision is a standard TTA mechanism. If the claimed results hold after the revisions below, the method would be a useful plug-and-play contribution to tubular structure segmentation under domain shift.

major comments (3)
  1. [Sec. 3.4, Eq. (9)-(10)] Stage 2 is supervised entirely by teacher pseudo-labels, yet the paper provides no measurement of pseudo-label correctness in the selected keypoints or pseudo-break regions. The confidence threshold tau = 0.5 does not guarantee accuracy under domain shift, and overconfident false-positive pixels can be selected and then reinforced by the cross-entropy term weighted by gamma in Eq. (10). This is load-bearing for the TopoHG contribution (Table 3: clDice increases from 62.20 to 66.61). I request a direct analysis: report pseudo-label precision/recall or agreement with ground truth at the selected keypoints and pseudo-break patches on the target-domain test sets, or provide an oracle-quality ablation (for example, comparing against a variant that uses ground-truth masks as pseudo-labels) to separate the effect of reliable supervision from the effect of the TopoHG operation. Without this, the claim that TopoHG improves topological continuity by correcting real breaks remains a plausible interpretation rather than an established one.
  2. [Abstract; Table 2] The advertised 'average improvement of 31.81% in clDice' is the absolute difference between Source Only and TopoTTA for the UNet baseline (Table 2: 42.19 to 74.00), not an average over prior TTA methods and not an average over both backbones. Against the strongest prior method (CoTTA), the average clDice gains are 4.66 points for UNet and 2.86 points for CS2Net; the average absolute gain over Source Only across both backbones is approximately 26.7 points. The claim as stated overstates the state-of-the-art margin and should be rewritten in percentage points with the reference baseline clearly identified.
  3. [Table 1; Table 2] The main results table is not numerically legible in the version under review. In the DRIVE to STARE block, the Source Only row reads '48.05 / 106.30', and several cells (105.20, 106.00, 298.74) are hard to assign to Dice, clDice, or Betti-error columns; the caption explains the '/' marker, but the column alignment makes it impossible to verify which values were averaged. This matters because Table 2 reports averages over ten datasets while at least three Source Only clDice entries are marked '/'; the paper never states how missing clDice values are handled in the average (available-case mean, imputation, or exclusion). Please regenerate Table 1 with unambiguous column separators, state the missing-value policy for computing averages, and recompute any headline numbers if the averaging rule changes the results.
minor comments (5)
  1. [Eq. (8)] The sentence preceding Eq. (8) reads 'we update only update pixels'; it should be 'we update only pixels'.
  2. [Sec. 4.1; Eq. (10); Sec. 3.4 Step 1] Eq. (10) introduces a weight gamma and Step 1 introduces a keypoint-selection coefficient, but the hyperparameter list in Sec. 4.1 only reports s, theta_bg, and the mask threshold; please give the values used for gamma and the keypoint coefficient in the main text.
  3. [Table C.3] Many p-values in Table C.3 are rendered as unreadable placeholders ('������'); please replace them with actual numeric values so the statistical significance claim can be checked.
  4. [Sec. 4.3] TopoHG is introduced as a three-step procedure, but Sec. 4.3 later refers to 'the above four steps'; please renumber consistently.
  5. [Supplementary Material, Sec. B.1] The code availability link is an anonymous placeholder; a stable release link should be provided if the paper is accepted.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: TopoTTA's central claims are evaluated against external ground-truth benchmarks and its self-supervised stages do not reduce, by equation or by definition, to the quantities they claim to improve.

full rationale

The paper's central claim—that TopoTTA improves Dice, clDice, and Betti errors across ten cross-domain datasets—is measured against ground-truth labels of external benchmarks (Sec. 4, Tables 1 and 2). No component is fitted to those labels. Stage 1 optimizes per-sample router parameters using entropy minimization (Eq. 6), and Stage 2 performs teacher-student consistency between predictions on TopoHG-edited inputs and teacher pseudo-labels (Eqs. 7-10), with fixed thresholds (tau=0.5, theta_bg=0.05). These are standard self-supervised TTA objectives, not predictions derived from target ground truth. Using the teacher's own outputs as pseudo-labels is the established CoTTA-style paradigm; it does not make the evaluation circular because the reported metrics compare final predictions with held-out ground truth, and no equation reduces to its own input. The paper's self-citations (e.g., refs. 14, 29-32, 47, 62, 77) are background/related-work citations and none is load-bearing: no uniqueness theorem, no smuggled ansatz, and no fitted constant is imported from them. The skeptic's concern about pseudo-label reliability is a robustness/correctness risk, not a circularity, and the paper acknowledges mitigation via confidence thresholding and foreground-ratio filtering in TopoHG. The 31.81% headline is an absolute-point gain over Source Only rather than over the best prior TTA, which is a framing issue but not a circular derivation. Under the stated rules—which require exhibiting a specific reduction such as Eq. X = Eq. Y by construction or a fitted parameter renamed as a prediction—no circular step can be identified, so the appropriate score is 0.

Assumptions & free parameters 5 free parameters · 4 assumptions · 2 invented entities

The central claim rests on standard TTA objectives (entropy minimization, pseudo-label consistency) and on task-specific hyperparameters chosen by hand. The algorithmic components TopoMDCs and TopoHG are introduced as invented entities with in-paper validation only. No physical constants or new conserved quantities are involved.

free parameters (5)
  • mask threshold = 0.5
    Binarization threshold used for all predictions; set to 0.5, standard but affects clDice calculation.
  • TopoHG window size s = 30
    Size of the foreground and background sliding windows in pseudo-break generation; tuned in Fig. C.3(a).
  • background foreground ratio threshold theta_bg = 0.05
    Upper limit on foreground pixels in the background window; tuned in Fig. C.3(d).
  • key point coefficient alpha = 0.002
    Controls the number of key points selected in TopoHG Step 1; tuned in Fig. C.3(c).
  • number of adaptation iterations = 6 (3 per stage)
    Chosen as a trade-off between performance and inference time; additional iterations give marginal gains (Table C.4).
assumptions (4)
  • domain assumption Entropy minimization is an effective proxy for adaptation at test time
    Stage 1 updates router parameters by minimizing prediction entropy (Eq. 6), a standard TTA objective but unproven for topology-aware routers.
  • domain assumption Teacher pseudo-labels on target images are trustworthy enough for supervision
    Stage 2 consistency loss (Eq. 9) relies on pseudo-labels from the teacher; if these are wrong, errors propagate.
  • domain assumption Low-frequency swapping between foreground and background patches yields valid hard samples
    TopoHG assumes the swapped patch retains foreground identity while creating a break (Eq. 7-8).
  • ad hoc to paper TopoMDCs can represent tubular topology despite being linear recombinations of vanilla convolutions
    The method postulates that directional difference terms improve topological representation; supported by ablations but not by theory.
invented entities (2)
  • Topological Meta Difference Convolutions (TopoMDCs)
    purpose: Replace vanilla convolutions in the encoder with direction-aware difference convolutions to enhance topology perception
    Validated only through in-paper experiments (Table 4, Fig. 4); no external or theoretical evidence.
  • Topology Hard sample Generation (TopoHG)
    purpose: Generate pseudo-break images by low-frequency swapping to create hard samples for consistency training
    Ablation shows benefit over blur, noise, and spatial swap, but evidence is limited to the paper's own benchmarks.

how reviews work

0 comments
Cite this review

Pith. "Pith review of TopoTTA: Topology-Enhanced Test-Time Adaptation for Tubular Structure Segmentation." pith.science (2026). https://pith.science/paper/CCRIYO34

@misc{pith2026250800442,
  author       = {Pith},
  title        = {Pith review of: TopoTTA: Topology-Enhanced Test-Time Adaptation for Tubular Structure Segmentation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/CCRIYO34}},
  note         = {Machine review of arXiv:2508.00442}
}
read the original abstract

Tubular structure segmentation (TSS) is important for various applications, such as hemodynamic analysis and route navigation. Despite significant progress in TSS, domain shifts remain a major challenge, leading to performance degradation in unseen target domains. Unlike other segmentation tasks, TSS is more sensitive to domain shifts, as changes in topological structures can compromise segmentation integrity, and variations in local features distinguishing foreground from background (e.g., texture and contrast) may further disrupt topological continuity. To address these challenges, we propose Topology-enhanced Test-Time Adaptation (TopoTTA), the first test-time adaptation framework designed specifically for TSS. TopoTTA consists of two stages: Stage 1 adapts models to cross-domain topological discrepancies using the proposed Topological Meta Difference Convolutions (TopoMDCs), which enhance topological representation without altering pre-trained parameters; Stage 2 improves topological continuity by a novel Topology Hard sample Generation (TopoHG) strategy and prediction alignment on hard samples with pseudo-labels in the generated pseudo-break regions. Extensive experiments across four scenarios and ten datasets demonstrate TopoTTA's effectiveness in handling topological distribution shifts, achieving an average improvement of 31.81% in clDice. TopoTTA also serves as a plug-and-play TTA solution for CNN-based TSS models.

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

77 extracted references · 76 canonical work pages

  1. [1]

    Source-Free Domain Adaptive Fundus Image Segmentation with Denoised Pseudo-Labeling

    Cheng Chen, Quande Liu, Yueming Jin, Qi Dou, and Pheng- Ann Heng. Source-Free Domain Adaptive Fundus Image Segmentation with Denoised Pseudo-Labeling. In Proceed- ings of the International Conference on Medical Image Com- puting and Computer Assisted Interventions, pages 225–235,

    work page

  2. [2]

    Mind Marginal Non-Crack Regions: Clustering-Inspired Representation Learning for Crack Segmentation

    Zhuangzhuang Chen, Zhuonan Lai, Jie Chen, and Jianqiang Li. Mind Marginal Non-Crack Regions: Clustering-Inspired Representation Learning for Crack Segmentation. In Pro- ceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pages 12698–12708, 2024. 3

    work page 2024

  3. [3]

    Each Test Image Deserves a Specific Prompt: Continual Test-Time Adaptation for 2D Medical Image Seg- mentation

    Ziyang Chen, Yongsheng Pan, Yiwen Ye, Mengkang Lu, and Yong Xia. Each Test Image Deserves a Specific Prompt: Continual Test-Time Adaptation for 2D Medical Image Seg- mentation. In Proceedings of the IEEE Conference on Com- puter Vision and Pattern Recognition , pages 11184–11193,

    work page

  4. [4]

    Automatic Road Detection and Centerline Extraction via Cascaded End-to- End Convolutional Neural Network

    Guangliang Cheng, Ying Wang, Shibiao Xu, Hongzhen Wang, Shiming Xiang, and Chunhong Pan. Automatic Road Detection and Centerline Extraction via Cascaded End-to- End Convolutional Neural Network. IEEE Transactions on Geoscience and Remote Sensing, 55(6):3322–3337, 2017. 6

    work page 2017

  5. [5]

    Compound Batch Normalization for Long- Tailed Image Classification

    Lechao Cheng, Chaowei Fang, Dingwen Zhang, Guanbin Li, and Gang Huang. Compound Batch Normalization for Long- Tailed Image Classification. In Proceedings of the ACM International Conference on Multimedia, pages 1925–1934,

    work page 1925

  6. [6]

    Joint Topology-Preserving and Feature-Refinement Network for Curvilinear Structure Segmentation

    Mingfei Cheng, Kaili Zhao, Xuhong Guo, Yajing Xu, and Jun Guo. Joint Topology-Preserving and Feature-Refinement Network for Curvilinear Structure Segmentation. In Pro- ceedings of the IEEE International Conference on Computer Vision, pages 7147–7156, 2021. 1

    work page 2021

  7. [7]

    Cooley and John W

    James W. Cooley and John W. Tukey. An Algorithm for the Machine Calculation of Complex Fourier Series. Mathemat- ics of Computation, 19(90):297–301, 1965. 5

    work page 1965

  8. [8]

    DeepGlobe 2018: A Challenge to Parse the Earth Through Satellite Images

    Ilke Demir, Krzysztof Koperski, David Lindenbaum, Guan Pang, Jing Huang, Saikat Basu, Forest Hughes, Devis Tuia, and Ramesh Raskar. DeepGlobe 2018: A Challenge to Parse the Earth Through Satellite Images. In IEEE Conference on Computer Vision and Pattern Recognition Workshop, pages 172–181, 2018. 6

    work page 2018

Show all 77 references
  1. [9]

    What Can Be Transferred: Unsupervised Domain Adaptation for Endoscopic Lesions Segmentation

    Jiahua Dong, Yang Cong, Gan Sun, Bineng Zhong, and Xi- aowei Xu. What Can Be Transferred: Unsupervised Domain Adaptation for Endoscopic Lesions Segmentation. In Pro- ceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pages 4023–4032, 2020. 1

    2020

  2. [10]

    Robust Mean Teacher for Continual and Gradual Test-Time Adapta- tion

    Mario D ¨obler, Robert A Marsden, and Bin Yang. Robust Mean Teacher for Continual and Gradual Test-Time Adapta- tion. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pages 7704–7714, 2023. 2, 3

    2023

  3. [11]

    Semi- Supervised Heterogeneous Domain Adaptation: Theory and Algorithms

    Zhen Fang, Jie Lu, Feng Liu, and Guangquan Zhang. Semi- Supervised Heterogeneous Domain Adaptation: Theory and Algorithms. IEEE Transactions on Pattern Analysis and Ma- chine Intelligence, 45(1):1087–1105, 2022. 1

    2022

  4. [12]

    An Ensemble Classification-based Approach Applied to Retinal Blood Vessel Segmentation

    Muhammad Moazam Fraz, Paolo Remagnino, Andreas Hoppe, Bunyarit Uyyanonvara, Alicja R Rudnicka, Christo- pher G Owen, and Sarah A Barman. An Ensemble Classification-based Approach Applied to Retinal Blood Vessel Segmentation. IEEE Transactions on Biomedical En- gineering, 59(9)...

    2012

  5. [13]

    Mohsen Ghafoorian, Alireza Mehrtash, Tina Kapur, Nico Karssemeijer, Elena Marchiori, Mehran Pesteie, Charles R. G. Guttmann, Frank-Erik de Leeuw, Clare M. Tempany, Bram van Ginneken, Andriy Fedorov, Purang Abolmaesumi, Bram Platel, and William M. Wells. Transfer Learning for D...

    2017

  6. [14]

    MorFormer: morphology- aware transformer for generalized pavement crack segmen- tation

    Xin Guo, Wenzhong Tang, Haoran Wang, Jiale Wang, Shuai Wang, Xiaolei Qu, and Xun Lin. MorFormer: morphology- aware transformer for generalized pavement crack segmen- tation. IEEE Transactions on Intelligent Transportation Sys- tems, 26(6):8219–8232, 2025. 2

    2025

  7. [15]

    Topology-Aware Uncertainty for Image Segmentation

    Saumya Gupta, Yikai Zhang, Xiaoling Hu, Prateek Prasanna, and Chao Chen. Topology-Aware Uncertainty for Image Segmentation. In Advances in the Neural Information Pro- cessing Systems, page 36, 2024. 1

    2024

  8. [16]

    Locating Blood Vessels in Retinal Images by Piece- wise Threshold Probing of a Matched Filter Response

    AD Hoover, Valentina Kouznetsova, and Michael Gold- baum. Locating Blood Vessels in Retinal Images by Piece- wise Threshold Probing of a Matched Filter Response. IEEE Transactions on Medical Imaging, 19(3):203–210, 2000. 6

    2000

  9. [17]

    Structure-Aware Image Segmentation with Ho- motopy Warping

    Xiaoling Hu. Structure-Aware Image Segmentation with Ho- motopy Warping. In Advances in the Neural Information Processing Systems, pages 24046–24059, 2022. 3

    2022

  10. [18]

    Topology-Preserving Deep Image Segmentation

    Xiaoling Hu, Fuxin Li, Dimitris Samaras, and Chao Chen. Topology-Preserving Deep Image Segmentation. In Ad- vances in the Neural Information Processing Systems, 2019. 6

    2019

  11. [19]

    Topology-Aware Segmentation Using Discrete Morse Theory

    Xiaoling Hu, Yusu Wang, Li Fuxin, Dimitris Samaras, and Chao Chen. Topology-Aware Segmentation Using Discrete Morse Theory. In Proceedings of the International Confer- ence on Learning Representations, 2021. 3

    2021

  12. [20]

    Learn- ing Probabilistic Topological Representations Using Dis- crete Morse Theory

    Xiaoling Hu, Dimitris Samaras, and Chao Chen. Learn- ing Probabilistic Topological Representations Using Dis- crete Morse Theory. In Proceedings of the International Conference on Learning Representations, 2022. 3

    2022

  13. [21]

    Representing Topological Self-Similarity Using Fractal Feature Maps for Accurate Segmentation of Tubular Structures

    Jiaxing Huang, Yanfeng Zhou, Yaoru Luo, Guole Liu, Heng Guo, and Ge Yang. Representing Topological Self-Similarity Using Fractal Feature Maps for Accurate Segmentation of Tubular Structures. In Proceedings of the European Confer- ence on Computer Vision, pages 143–160, 2025. 1, 3

    2025

  14. [22]

    Single Image Test-Time Adaptation for Segmenta- tion

    Klara Janouskova, Tamir Shor, Chaim Baskin, and Jiri Matas. Single Image Test-Time Adaptation for Segmenta- tion. Transactions on Machine Learning Research, 2024. 2

    2024

  15. [23]

    Skeleton Recall Loss for Connectiv- ity Conserving and Resource Efficient Segmentation of Thin Tubular Structures

    Yannick Kirchhoff, Maximilian R Rokuss, Saikat Roy, Balint Kovacs, Constantin Ulrich, Tassilo Wald, Maximil- ian Zenk, Philipp V ollmuth, Jens Kleesiek, Fabian Isensee, and Klaus Maier-Hein. Skeleton Recall Loss for Connectiv- ity Conserving and Resource Efficient Segmentation...

    2024

  16. [24]

    CS2-Net: Deep learning Segmentation of Curvilinear Structures in Medical Imaging

    Mou Lei, Yitian Zhao, Huazhu Fu, Yonghuai Liu, Jun Cheng, Yalin Zheng, Pan Su, Jianlong Yang, Li Chen, Alejandro F Frangi, Masahiro Akiba, and Jiang Liu. CS2-Net: Deep learning Segmentation of Curvilinear Structures in Medical Imaging. Medical Image Analysis, 67:101874, 2021. 1, 6

    2021

  17. [25]

    Enriching Information and Preserving Semantic Consistency in Expanding Curvi- linear Object Segmentation Datasets

    Qin Lei, Jiang Zhong, and Qizhu Dai. Enriching Information and Preserving Semantic Consistency in Expanding Curvi- linear Object Segmentation Datasets. In Proceedings of the European Conference on Computer Vision, pages 233–250,

  18. [26]

    OCTA-500: A Retinal Dataset for Optical Coherence Tomography Angiography Study

    Mingchao Li, Kun Huang, Qiuzhuo Xu, Jiadong Yang, Yuhan Zhang, Zexuan Ji, Keren Xie, Songtao Yuan, Qinghuai Liu, and Qiang Chen. OCTA-500: A Retinal Dataset for Optical Coherence Tomography Angiography Study. Medical Image Analysis, 93:103092, 2024. 6

    2024

  19. [27]

    A Comprehensive Sur- vey on Test-Time Adaptation Under Distribution Shifts

    Jian Liang, Ran He, and Tieniu Tan. A Comprehensive Sur- vey on Test-Time Adaptation Under Distribution Shifts. In- ternational Journal of Computer Vision , pages 1–34, 2024. 2

    2024

  20. [28]

    Wong, and Xiaoying Tang

    Li Lin, Linkai Peng, Huaqing He, Pujin Cheng, Jiewei Wu, Kenneth K.Y . Wong, and Xiaoying Tang. YoloCurvSeg: You Only Label One Noisy Skeleton for Vessel-Style Curvilin- ear Structure Segmentation. Medical Image Analysis , 90: 102937, 2023. 3

    2023

  21. [29]

    Xun Lin, Shuai Wang, Rizhao Cai, Yizhong Liu, Ying Fu, Wenzhong Tang, Zitong Yu, and Alex C. Kot. Suppress and rebalance: Towards generalized multi-modal face anti- spoofing. In Proceedings of the IEEE Conference on Com- puter Vision and Pattern Recognition, pages 211–221, 2024. 3

    2024

  22. [30]

    Xun Lin, Yi Yu, Song Xia, Jue Jiang, Haoran Wang, Zitong Yu, Yizhong Liu, Ying Fu, Shuai Wang, Wen- zhong Tang, and Alex C. Kot. Safeguarding medical im- age segmentation datasets against unauthorized training via contour- and texture-aware perturbations. arXiv preprint arXiv:...

    2024 arXiv

  23. [31]

    Xun Lin, Yi Yu, Zitong Yu, Ruohan Meng, Jiale Zhou, Ajian Liu, Yizhong Liu, Shuai Wang, Wenzhong Tang, Zhen Lei, and Alex C. Kot. Hidemia: Hidden wavelet mining for privacy-enhancing medical image analysis. In Proceedings of the ACM International Conference on Multimedia , pag...

    2024

  24. [32]

    Reliable and balanced transfer learning for generalized mul- timodal face anti-spoofing

    Xun Lin, Ajian Liu, Zitong Yu, Rizhao Cai, Shuai Wang, Yi Yu, Jun Wan, Zhen Lei, Xiaochun Cao, and Alex Kot. Reliable and balanced transfer learning for generalized mul- timodal face anti-spoofing. IEEE Transactions on Pattern Analysis and Machine Intelligence, pages 1–18, 2025. 3

    2025

  25. [33]

    Crackformer: Transformer Network for Fine-Grained Crack Detection

    Huajun Liu, Xiangyu Miao, Christoph Mertz, Chengzhong Xu, and Hui Kong. Crackformer: Transformer Network for Fine-Grained Crack Detection. In Proceedings of the IEEE International Conference on Computer Vision, pages 3783– 3792, 2021. 2

    2021

  26. [34]

    FedDG: Federated Domain Generalization on Medical Image Segmentation via Episodic Learning in Con- tinuous Frequency Space

    Quande Liu, Cheng Chen, Jing Qin, Qi Dou, and Pheng- Ann Heng. FedDG: Federated Domain Generalization on Medical Image Segmentation via Episodic Learning in Con- tinuous Frequency Space. In Proceedings of the IEEE Con- ference on Computer Vision and Pattern Recognition, pages ...

    2021

  27. [35]

    Source-Free Do- main Adaptation for Semantic Segmentation

    Yuang Liu, Wei Zhang, and Jun Wang. Source-Free Do- main Adaptation for Semantic Segmentation. In Proceed- ings of the IEEE Conference on Computer Vision and Pattern Recognition, pages 1215–1224, 2021. 1

    2021

  28. [36]

    AADG: Automatic Augmentation for Domain Generalization on Retinal Image Segmentation

    Junyan Lyu, Yiqi Zhang, Yijin Huang, Li Lin, Pujin Cheng, and Xiaoying Tang. AADG: Automatic Augmentation for Domain Generalization on Retinal Image Segmentation. IEEE Transactions on Medical Imaging, 41(12):3699–3711,

  29. [37]

    ROSE: A Retinal OCT-Angiography Vessel Segmentation Dataset and New Model

    Yuhui Ma, Huaying Hao, Jianyang Xie, Huazhu Fu, Jiong Zhang, Jianlong Yang, Zhen Wang, Jiang Liu, Yalin Zheng, and Yitian Zhao. ROSE: A Retinal OCT-Angiography Vessel Segmentation Dataset and New Model. IEEE Transactions on Medical Imaging, 40(3):928–939, 2020. 1, 6

    2020

  30. [38]

    The Norm Must Go On: Dynamic Unsu- pervised Domain Adaptation by Normalization

    M Jehanzeb Mirza, Jakub Micorek, Horst Possegger, and Horst Bischof. The Norm Must Go On: Dynamic Unsu- pervised Domain Adaptation by Normalization. In Proceed- ings of the IEEE Conference on Computer Vision and Pattern Recognition, pages 14765–14775, 2022. 3

    2022

  31. [39]

    Machine Learning for Aerial Image La- beling

    V olodymyr Mnih. Machine Learning for Aerial Image La- beling. University of Toronto, Canada, 2013. 6

    2013

  32. [40]

    A Tuan Nguyen, Thanh Nguyen-Tang, Ser-Nam Lim, and Philip H.S. Torr. Tipi: Test Time Adaptation with Transfor- mation Invariance. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pages 24162– 24171, 2023. 2

    2023

  33. [41]

    Distribution-Aware Continual Test-Time Adaptation for Se- mantic Segmentation

    Jiayi Ni, Senqiao Yang, Ran Xu, Jiaming Liu, Xiaoqi Li, Wenyu Jiao, Zehui Chen, Yi Liu, and Shanghang Zhang. Distribution-Aware Continual Test-Time Adaptation for Se- mantic Segmentation. In Proceedings of International Con- ference on Robotics and Automation , pages 3044–3050,

  34. [42]

    MADAv2: Advanced Multi-Anchor based Active Domain Adaptation Segmentation

    Munan Ning, Donghuan Lu, Yujia Xie, Dongdong Chen, Dong Wei, Yefeng Zheng, Yonghong Tian, Shuicheng Yan, and Li Yuan. MADAv2: Advanced Multi-Anchor based Active Domain Adaptation Segmentation. IEEE Transac- tions on Pattern Analysis and Machine Intelligence, 45(11): 13553–1356...

    2023

  35. [43]

    Towards Stable Test-Time Adaptation in Dynamic Wild World

    Shuaicheng Niu, Jiaxiang Wu, Yifan Zhang, Zhiquan Wen, Yaofo Chen, Peilin Zhao, and Mingkui Tan. Towards Stable Test-Time Adaptation in Dynamic Wild World. In Proceed- ings of the International Conference on Learning Represen- tations, 2023. 2, 3, 6, 4, 5

    2023

  36. [44]

    MedBN: Robust Test-Time Adapta- tion Against Malicious Test Samples

    Hyejin Park, Jeongyeon Hwang, Sunung Mun, Sangdon Park, and Jungseul Ok. MedBN: Robust Test-Time Adapta- tion Against Malicious Test Samples. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recogni- tion, pages 5997–6007, 2024. 3, 6, 4, 5

    2024

  37. [45]

    A Survey on Curvilinear Object Segmentation in Multiple Applications

    Bibiloni Pedro, Gonz ´alez-Hidalgo M, and Massanet Se- basti`a. A Survey on Curvilinear Object Segmentation in Multiple Applications. Pattern Recognition, 60:949–970,

  38. [46]

    Dynamic Snake Convolution Based on Topo- logical Geometric Constraints for Tubular Structure Segmen- tation

    Yaolei Qi, Yuting He, Xiaoming Qi, Yuan Zhang, and Guanyu Yang. Dynamic Snake Convolution Based on Topo- logical Geometric Constraints for Tubular Structure Segmen- tation. In Proceedings of the IEEE International Conference on Computer Vision, pages 6070–6079, 2023. 1, 2, 6

    2023

  39. [47]

    Eh- former: Regional easy-hard-aware transformer for breast le- sion segmentation in ultrasound images.Information Fusion, 109:102430, 2024

    Xiaolei Qu, Jiale Zhou, Jue Jiang, Wenhan Wang, Haoran Wang, Shuai Wang, Wenzhong Tang, and Xun Lin. Eh- former: Regional easy-hard-aware transformer for breast le- sion segmentation in ultrasound images.Information Fusion, 109:102430, 2024. 1

    2024

  40. [48]

    U- Net: Convolutional Networks for Biomedical Image Seg- mentation

    Olaf Ronneberger, Philipp Fischer, and Thomas Brox. U- Net: Convolutional Networks for Biomedical Image Seg- mentation. In Proceedings of the International Conference on Medical Image Computing and Computer Assisted Inter- ventions, pages 234–241, 2015. 6

    2015

  41. [49]

    Align Your Prompts: Test- Time Prompting with Distribution Alignment for Zero-Shot Generalization

    Jameel Hassan Abdul Samadh, Hanan Gani, Noor Hazim Hussein, Muhammad Uzair Khattak, Muzammal Naseer, Fa- had Khan, and Salman Khan. Align Your Prompts: Test- Time Prompting with Distribution Alignment for Zero-Shot Generalization. In Advances in the Neural Information Pro- ces...

    2023

  42. [50]

    Local Intensity Order Transformation for Robust Curvilinear Object Segmentation

    Tianyi Shi, Nicolas Boutry, Yongchao Xu, and Thierry G´eraud. Local Intensity Order Transformation for Robust Curvilinear Object Segmentation. IEEE Transactions on Im- age Processing, 31:2557–2569, 2022. 1, 2

    2022

  43. [51]

    FreeCOS: Self-Supervised Learning from Fractals and Unla- beled Images for Curvilinear Object Segmentation

    Tianyi Shi, Xiaohuan Ding, Liang Zhang, and Xin Yang. FreeCOS: Self-Supervised Learning from Fractals and Unla- beled Images for Curvilinear Object Segmentation. In Pro- ceedings of the IEEE International Conference on Computer Vision, pages 876–886, 2023. 3

    2023

  44. [52]

    Pluim, Ulrich Bauer, and Bjoern H Menze

    Suprosanna Shit, Johannes C Paetzold, Anjany Sekuboyina, Ivan Ezhov, Alexander Unger, Andrey Zhylka, Josien P.W. Pluim, Ulrich Bauer, and Bjoern H Menze. clDice: A Novel Topology-Preserving Loss Function for Tubular Structure Segmentation. In Proceedings of the IEEE Conference...

    2021

  45. [53]

    ECoTTA: Memory-Efficient Continual Test-Time Adapta- tion via Self-Distilled Regularization

    Junha Song, Jungsoo Lee, In So Kweon, and Sungha Choi. ECoTTA: Memory-Efficient Continual Test-Time Adapta- tion via Self-Distilled Regularization. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recogni- tion, pages 11920–11929, 2023. 2, 3

    2023

  46. [54]

    Ridge-based Vessel Segmentation in Color Images of the Retina

    Joes Staal, Michael D Abr `amoff, Meindert Niemeijer, Max A Viergever, and Bram Van Ginneken. Ridge-based Vessel Segmentation in Color Images of the Retina. IEEE Trans- actions on Medical Imaging, 23(4):501–509, 2004. 6

    2004

  47. [55]

    Reti- nal Vessel Segmentation with Skeletal Prior and Contrastive Loss

    Yubo Tan, Kaifu Yang, Shixuan Zhao, and Yongjie Li. Reti- nal Vessel Segmentation with Skeletal Prior and Contrastive Loss. IEEE Transactions on Medical Imaging, 41(9):2238– 2251, 2022. 1

    2022

  48. [56]

    TeSLA: Test-Time Self-Learning with Automatic Adversarial Augmentation

    Devavrat Tomar, Guillaume Vray, Behzad Bozorgtabar, and Jean-Philippe Thiran. TeSLA: Test-Time Self-Learning with Automatic Adversarial Augmentation. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recogni- tion, pages 20341–20350, 2023. 3

    2023

  49. [57]

    Tubular Structure Segmentation Using Spatial Fully Connected Network with Radial Distance Loss for 3D Med- ical Images

    Chenglong Wang, Yuichiro Hayashi, Masahiro Oda, Hayato Itoh, Takayuki Kitasaka, Alejandro F Frangi, and Kensaku Mori. Tubular Structure Segmentation Using Spatial Fully Connected Network with Radial Distance Loss for 3D Med- ical Images. In Proceedings of the International Con...

    2019

  50. [58]

    TENT: Fully Test-Time Adap- tation by Entropy Minimization

    Dequan Wang, Evan Shelhamer, Shaoteng Liu, Bruno Ol- shausen, and Trevor Darrell. TENT: Fully Test-Time Adap- tation by Entropy Minimization. Proceedings of the Interna- tional Conference on Learning Representations, 2021. 1, 2, 3, 5, 6, 4

    2021

  51. [59]

    PointScatter: Point Set Representa- tion for Tubular Structure Extraction

    Dong Wang, Zhao Zhang, Ziwei Zhao, Yuhang Liu, Yihong Chen, and Liwei Wang. PointScatter: Point Set Representa- tion for Tubular Structure Extraction. In Proceedings of the European Conference on Computer Vision, pages 366–383,

  52. [60]

    Con- tinual Test-Time Domain Adaptation

    Qin Wang, Olga Fink, Luc Van Gool, and Dengxin Dai. Con- tinual Test-Time Domain Adaptation. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recogni- tion, pages 7201–7211, 2022. 1, 2, 3, 6, 4, 5

    2022

  53. [61]

    Dynamically Instance- Guided Adaptation: A Backward-Free Approach for Test- Time Domain Adaptive Semantic Segmentation

    Wei Wang, Zhun Zhong, Weijie Wang, Xi Chen, Charles Ling, Boyu Wang, and Nicu Sebe. Dynamically Instance- Guided Adaptation: A Backward-Free Approach for Test- Time Domain Adaptive Semantic Segmentation. InProceed- ings of the IEEE Conference on Computer Vision and Pattern Rec...

    2023

  54. [62]

    Interactively fusing global and local features for benign and malignant classification of breast ultrasound images

    Wenhan Wang, Jiale Zhou, Jin Zhao, Xun Lin, Yan Zhang, Shan Lu, Wanchen Zhao, Shuai Wang, Wenzhong Tang, and Xiaolei Qu. Interactively fusing global and local features for benign and malignant classification of breast ultrasound images. Ultrasound in Medicine & Biology, 51(3):525–534,

  55. [63]

    Fishman, and Alan L

    Yan Wang, Xu Wei, Fengze Liu, Jieneng Chen, Yuyin Zhou, Wei Shen, Elliot K. Fishman, and Alan L. Yuille. Deep Dis- tance Transform for Tubular Structure Segmentation in CT Scans. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2020. 1, 3

    2020

  56. [64]

    DDU- Net: Dual-Decoder-U-Net for Road Extraction Using High- Resolution Remote Sensing Images

    Ying Wang, Yuexing Peng, Wei Li, George C Alexandropou- los, Junchuan Yu, Daqing Ge, and Wei Xiang. DDU- Net: Dual-Decoder-U-Net for Road Extraction Using High- Resolution Remote Sensing Images. IEEE Transactions on Geoscience and Remote Sensing, 60:1–12, 2022. 1

    2022

  57. [65]

    Deep Closing: Enhancing Topological Connectiv- ity in Medical Tubular Segmentation

    Qian Wu, Yufei Chen, Wei Liu, Xiaodong Yue, and Xiahai Zhuang. Deep Closing: Enhancing Topological Connectiv- ity in Medical Tubular Segmentation. IEEE Transactions on Medical Imaging, 43(11):3990–4003, 2024. 1, 6

    2024

  58. [66]

    A Topological-Attention ConvLSTM Network and Its Appli- cation to EM Images

    Jiaqi Yang, Xiaoling Hu, Chao Chen, and Chialing Tsai. A Topological-Attention ConvLSTM Network and Its Appli- cation to EM Images. In Proceedings of the International Conference on Medical Image Computing and Computer As- sisted Interventions, pages 217–228, 2021. 2

    2021

  59. [67]

    FDA: Fourier Domain Adaptation for Semantic Segmentation

    Yanchao Yang and Stefano Soatto. FDA: Fourier Domain Adaptation for Semantic Segmentation. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pages 4085–4095, 2020. 1

    2020

  60. [68]

    Directional Connectivity-based Segmentation of Medical Images

    Ziyun Yang and Sina Farsiu. Directional Connectivity-based Segmentation of Medical Images. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recog- nition, pages 11525–11535, 2023. 2, 6

    2023

  61. [69]

    Chang’an Yi, Haotian Chen, Yifan Zhang, Yonghui Xu, Yan Zhou, and Lizhen Cui. From Question to Exploration: Can Classic Test-Time Adaptation Strategies Be Effectively Applied in Semantic Segmentation? In Proceedings of the ACM International Conference on Multimedia , pages 100...

    2024

  62. [70]

    Searching Central Difference Convolutional Networks for Face Anti-Spoofing

    Zitong Yu, Chenxu Zhao, Zezheng Wang, Yunxiao Qin, Zhuo Su, Xiaobai Li, Feng Zhou, and Guoying Zhao. Searching Central Difference Convolutional Networks for Face Anti-Spoofing. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition , pages 5295– 5305,...

    2020

  63. [71]

    Robust Test- Time Adaptation in Dynamic Scenarios

    Longhui Yuan, Binhui Xie, and Shuang Li. Robust Test- Time Adaptation in Dynamic Scenarios. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pages 15922–15932, 2023. 2

    2023

  64. [72]

    Do- mainAdaptor: A Novel Approach to Test-Time Adaptation

    Jian Zhang, Lei Qi, Yinghuan Shi, and Yang Gao. Do- mainAdaptor: A Novel Approach to Test-Time Adaptation. In Proceedings of the IEEE International Conference on Computer Vision, pages 18971–18981, 2023. 2, 3, 6, 4, 5

    2023

  65. [73]

    Stagewise Unsupervised Domain Adaptation with Adversar- ial Self-Training for Road Segmentation of Remote-Sensing Images

    Lefei Zhang, Meng Lan, Jing Zhang, and Dacheng Tao. Stagewise Unsupervised Domain Adaptation with Adversar- ial Self-Training for Road Segmentation of Remote-Sensing Images. IEEE Transactions on Geoscience and Remote Sens- ing, 60:1–13, 2021. 1

    2021

  66. [74]

    TestFit: A Plug-and-Play One-Pass Test Time Method for Medical Image Segmentation

    Yizhe Zhang, Tao Zhou, Yuhui Tao, Shuo Wang, Ye Wu, Benyuan Liu, Pengfei Gu, Qiang Chen, and Danny Z Chen. TestFit: A Plug-and-Play One-Pass Test Time Method for Medical Image Segmentation. Medical Image Analysis, 92: 103069, 2024. 2

    2024

  67. [75]

    DELTA: Degradation-Free Fully Test-Time Adaptation

    Bowen Zhao, Chen Chen, and Shu-Tao Xia. DELTA: Degradation-Free Fully Test-Time Adaptation. In Proceed- ings of the International Conference on Learning Represen- tations, 2023. 3

    2023

  68. [76]

    Synthesizing Retinal and Neuronal Images with Generative Adversarial Nets

    He Zhao, Huiqi Li, Sebastian Maurer-Stroh, and Li Cheng. Synthesizing Retinal and Neuronal Images with Generative Adversarial Nets. Medical Image Analysis, 49:14–26, 2018. 6

    2018

  69. [77]

    Source Only

    Jiale Zhou, Zuoxun Hou, Hongyan Lu, Wenhan Wang, Wanchen Zhao, Zenan Wang, Dezhi Zheng, Shuai Wang, Wenzhong Tang, and Xiaolei Qu. A deep supervised trans- former u-shaped full-resolution residual network for the seg- mentation of breast ultrasound image. Medical Physics, 50 (...

    2023

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.