Pith. sign in

REVIEW 2 major objections 7 minor 76 references

Widest-Path Reachability Fields for Connectivity-Preserving Slender Structure Segmentation

T0 review · 2 major / 7 minor · reviewed 2026-07-09 · glm-5.2

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

Pith's one-line read Max-Min algebra fixes broken vessel segmentation by targeting

desk verdict Solid practical method with a real gap in the ablation logic — the Max-Min operator itself is never isolated from alternatives. read the letter →

arxiv 2607.07123 v1 pith:P5E4UPU6 submitted 2026-07-08 cs.CV cs.SYeess.SY

Youcheng Zong , Runda Jia , Minxuan Hu , Weilan Su , Dakuo He This is my paper · ORCID
classification cs.CVcs.SYeess.SY
keywords topologicalgradientstarvationMax-Minalgebrawidestpathproblemconnectivity-preservingsegmentationroutingdifferentiabledynamicprogrammingslenderstructurebottleneckpixels
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 identifies a systematic failure mode in segmenting thin curvilinear structures—retinal vessels, cracks, roads—where standard pixel-wise losses like Dice and BCE distribute gradient updates uniformly across all pixels, yet actual connectivity depends on a vanishingly small number of bottleneck pixels. Because these bottleneck pixels are so sparse, their aggregate gradient contribution is negligible compared to thick structures and background, so the network never learns to repair the breaks that fragment the network. The authors call this Topological Gradient Starvation (TGS). Their solution, Widest-Path Reachability Fields (WPRF), replaces sum-based gradient aggregation with Max-Min algebra: a path's strength equals its weakest edge, so backpropagation through the min operator concentrates gradients exclusively on the current bottleneck edge, and once that edge strengthens, the max operator shifts focus to the next weakest link. This produces a seek-and-repair dynamic that progressively fixes connectivity breaks without any inference-time overhead. The method is implemented as a plug-and-play training loss on top of any segmentation backbone, using differentiable dynamic programming over a skeleton-derived support graph, and improves connectivity metrics in 47 of 54 architecture–dataset combinations tested.

What carries the argument

Differentiable Max-Min dynamic programming over a skeleton-derived support graph. For each source node, the algorithm iteratively propagates reachability values using the recurrence r_{t+1}(v) = max(r_t(v), max_{p} min(r_t(p), w_{pv})), where w_{pv} is the learned edge weight between adjacent nodes. The min operator retains only the bottleneck edge's gradient per candidate path; the max operator selects the strongest path. Ties return a valid subgradient. The support graph is restricted to the ground-truth skeleton domain V* to prevent background shortcuts. Three loss terms are combined: a bottleneck-aware segmentation loss that upweights thin structures via inverse skeleton radius, a local

What would settle it

If replacing the Max-Min operator with a sum-based or product-based reachability formulation on the same domain-restricted support graph produced equivalent clDice improvements, the bottleneck-dominated gradient routing mechanism would not be the operative cause—instead, domain restriction and multi-scale supervision alone would explain the gains.

Watch

Extended reading notes

Core claim

The central claim is that connectivity failures in slender-structure segmentation stem not from inadequate model capacity or training data, but from an algebraic property of standard loss functions: sum-based operators spread gradients uniformly, starving the sparse bottleneck pixels that actually determine topological continuity. By replacing the sum operator with a Max-Min operator—where a path's reachability is defined by its weakest edge and the gradient flows only through that bottleneck—training automatically redirects learning signal to the pixels that matter most for connectivity. The differentiable k-step dynamic programming implementation on a domain-restricted support graph makes

Load-bearing premise

The method assumes that a deterministic skeletonization operator applied to the ground-truth binary mask faithfully captures the connectivity structure the network should learn, and that supervising reachability on this coarsened skeleton graph transfers to the full-resolution pixel mask used at inference. If the skeletonization introduces artifacts or the stride-based coarsening loses critical bottleneck topology, the Max-Min gradient routing targets may not align with real

Editorial extensions

If this is right

  • Any segmentation task where topology matters more than pixel overlap—neuron tracing, road network extraction, crack detection—could benefit from replacing or augmenting sum-based losses with Max-Min reachability objectives, since the gradient routing mechanism is architecture-agnostic and requires no skeleton labels beyond what is derivable from binary masks.
  • The seek-and-repair gradient pattern suggests a natural curriculum: as bottlenecks are progressively strengthened, the network automatically shifts focus, which could reduce or eliminate the need for multi-stage training pipelines or hard-example mining heuristics currently used in vessel and road segmentation.
  • The domain-restriction strategy—constraining graph propagation to skeleton-derived support rather than full foreground occupancy—demonstrates that preventing background shortcuts is as important as the Max-Min operator itself, which has implications for any graph-based learning signal where spurious paths could create shortcuts.
  • The identification of TGS as a structural property of sum-based losses, rather than a data or capacity problem, reframes connectivity preservation as an optimization geometry question that could be addressed at the loss-function level for many beyond-segmentation tasks involving graph-structured outputs.

Reading between the lines

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

  • The Max-Min gradient routing principle may extend to other domains where a sparse subset of elements determines global property correctness—e.g., chain-of-thought reasoning where one weak logical link breaks an argument, or program synthesis where a single incorrect operation invalidates output—suggesting that bottleneck-dominated gradient flow is a general optimization principle beyond pixel segm
  • The 2–3× training time overhead from Max-Min propagation could potentially be reduced by sparse or approximate variants that only propagate through candidate bottleneck regions identified by low-confidence edges, rather than full k-step DP over all nodes, which would make the approach practical for 3D volumetric data where the current formulation may be prohibitively expensive.
  • If the skeletonization operator Phi_px introduces systematic biases—e.g., preferentially retaining certain vessel calibers or missing sub-pixel connections—the Max-Min objective would faithfully optimize the wrong connectivity targets, suggesting that sensitivity to skeletonization quality is the critical failure mode that would distinguish between TGS as a fundamental phenomenon versus an artifac
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

2 major / 7 minor

Summary. This paper introduces Widest-Path Reachability Fields (WPRF), a plug-and-play training module for connectivity-preserving segmentation of slender curvilinear structures (vessels, cracks, roads). The authors identify Topological Gradient Starvation (TGS)—the phenomenon where sparse connectivity-critical bottleneck pixels receive insufficient gradients under standard sum-based pixel-wise losses—and propose a differentiable Max-Min reachability objective on a domain-restricted support graph to redirect gradient flow to these bottlenecks. The method is evaluated across nine architectures and six datasets (including a newly introduced oral microvessel dataset, OMVIS) with fixed hyperparameters, showing clDice improvements in 47 of 54 method-dataset pairs. The experimental design is thorough: multiple backbone families, ablations isolating each loss component, gradient routing verification (Fig. 7), and hyperparameter sensitivity analysis.

Significance. The paper addresses a well-known and practically important problem: topological breaks in curvilinear structure segmentation. The TGS formalization provides a clear diagnostic framing, and the Max-Min gradient routing mechanism is a principled solution grounded in classical graph theory (widest-path problem). Key strengths include: (1) the gradient routing verification in Fig. 7, which directly visualizes the shift of gradient energy from thick to thin structures; (2) the breadth of the experimental validation—nine architectures, six datasets, fixed hyperparameters, 9 runs per configuration; (3) the ablation in Table IX isolating edge supervision, reachability supervision, domain restriction, and multi-scale propagation; (4) the plug-and-play, backbone-agnostic design with no inference overhead (Table X). The newly introduced OMVIS dataset with expert annotation and inter-observer agreement reporting is a useful community contribution. Code is stated to be available.

major comments (2)
  1. The central novelty claim is that Max-Min algebra specifically induces bottleneck-dominated gradient routing (Section III-C, Fig. 3), distinguishing WPRF from prior sum-based and product-based methods (Section II). However, the ablation in Table IX only compares against pixel-level baselines (BCE, Focal, BCE+clDice). It never replaces the Max-Min DP in Eq. (9) with an alternative reachability formulation—e.g., a sum-based additive path cost or a product-based affinity chain—on the same domain-restricted support graph V* with the same multi-scale sampling, edge loss, and BA segmentation term. Without this comparison, the clDice gains cannot be attributed to the Max-Min operator specifically rather than to the general framework of graph-based reachability supervision + domain restriction + multi-scale propagation. The theoretical argument in Section III-C is plausible but not self-evident:
  2. Dice drops are observed in several configurations (e.g., Swin-UNet on OCTA-500 6mm: 0.801→0.775 in Table V; Swin-UNet on OCTA-500 3mm: 0.828→0.798 in Table IV; CS-Net on OCTA-500 3mm: 0.796→0.788 in Table IV). The paper acknowledges a 'slight Dice drop' in the ablation (Section IV-E) but does not systematically analyze when or why Dice degrades. For a method claiming to preserve connectivity 'without sacrificing region overlap' (Section IV-C), the conditions under which pixel-level overlap significantly decreases should be characterized. Is this a trade-off inherent to the bottleneck-aware weighting (Eq. 12), or does the reachability loss actively suppress certain foreground regions? A brief analysis of the Dice-clDice trade-off space would strengthen the paper.
minor comments (7)
  1. Section III-B, Eq. (2): the deterministic skeletonization operator Phi_px is described as 'closing (3x3) + Zhang-Suen thinning' in Table I but the text mentions 'lightweight morphological preprocessing to fill small gaps.' The sensitivity of the support graph V* to the choice of skeletonization algorithm is not discussed. A brief note on robustness to skeletonization quality would strengthen the paper.
  2. Table I states 'Augmentation: None.' For a method evaluated across six datasets with varying characteristics, the absence of any data augmentation is unusual and may disadvantage baselines. A brief justification for this choice would be helpful.
  3. Section IV-F.3, Table X: training time increases by 2-3x with WPRF (e.g., UNet on DRIVE: 9.7→79.7 ms/iter). While the paper notes this is training-only overhead, the practical implications for large-scale experiments are not discussed. The paper mentions 'sparse Max-Min propagation' as future work but does not quantify how much overhead could be reduced.
  4. Fig. 4: the y-axis label 'ΔclDice (pp)' and the markers are clear, but the 'dataset min-max' whiskers overlap with individual backbone markers in some cases, making it difficult to distinguish per-backbone values from the range. Consider jittering or using a different visual encoding.
  5. Section III-D, Eq. (10): the positive pair distance criterion dist ∈ [ceil(k/2), k] is motivated as excluding 'trivially reachable pairs,' but the choice of ceil(k/2) as the lower bound is not justified. Is this sensitive to the choice of lower bound?
  6. The paper introduces the term 'Bottleneck-Aware Balanced Hard-Negative BCE' for L_seg (Section III-D) but the 'bottleneck-aware' aspect refers to the weight W-bar(x) based on skeleton radius (Eq. 12), which is conceptually distinct from the Max-Min bottleneck routing in L_reach. The shared use of 'bottleneck' terminology for both the pixel-level reweighting and the graph-level Max-Min routing may cause confusion. Clarifying the relationship between the two mechanisms would help.
  7. Reference formatting: some entries have inconsistent capitalization and venue abbreviations (e.g., 'arXiv preprint' vs. full conference names). A pass through the reference list for consistency would improve presentation.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity found; derivation is self-contained against external benchmarks and classical graph theory

full rationale

The paper's derivation chain is not circular. (1) The TGS concept is an analogy to gradient starvation in classification (ref 18, Pezeshki et al., NeurIPS 2021 — external), not a self-cited result. (2) The Max-Min widest-path formulation derives from classical graph theory (refs 59–60, Smith 1993; Udupa & Samarasekera 1996 — external), and the k-step DP in Eq. (9) is a standard Max-Min semiring computation. (3) The gradient routing property (Section III-C) is an inherent mathematical property of min/max operators, not defined in terms of the paper's own outputs. (4) The support graph V* (Eq. 2) is constructed from GT masks via deterministic operators, not from predictions. (5) The bottleneck-aware weight (Eq. 12) is derived from skeleton radius of GT, not from the model's own predictions. (6) Pair labels for reachability supervision (Eq. 10) come from GT connected-component structure, not from predicted affinities. (7) Self-citations (refs 9, 24, 33) are tangential — a SAM2 case study, a manufacturing random forest, and a meta-contrastive learning model — none are invoked as load-bearing mathematical results or uniqueness theorems. (8) The OMVIS dataset is independently collected with ethical approval and inter-observer validation. (9) Experimental validation uses external datasets (DRIVE, OCTA-500, DeepCrack, Massachusetts Roads) and standard metrics (Dice, clDice). The skeptic's concern about Max-Min not being ablated against sum/product alternatives on the same graph is a valid experimental completeness concern, but it is a correctness/evaluation risk, not circularity — the paper does not define its prediction in terms of its own inputs or self-cited unverified results.

Assumptions & free parameters 7 free parameters · 4 assumptions · 3 invented entities

The method has seven free parameters, all fixed across datasets without per-dataset tuning. The axioms are domain assumptions standard in curvilinear structure segmentation (threshold connectivity, skeleton faithfulness, bottleneck criticality). No ad-hoc-to-paper axioms are introduced. The invented entities (TGS, WPRF, OMVIS) all have independent evidence or falsifiable handles.

free parameters (7)
  • lambda_edge = 1.0
    Loss weight for edge loss term, fixed across all datasets (Table I).
  • lambda_reach = 1.0
    Loss weight for reachability loss term, fixed across all datasets (Table I).
  • graph stride s = 4
    Controls coarsening from pixel grid to graph grid; chosen to balance accuracy and efficiency (Section IV-F).
  • N_s (sources per image) = 32
    Number of source nodes sampled per scale for reachability supervision (Table I).
  • multi-scale steps K = {1,2,4,8,16}
    Set of k values for multi-scale reachability; gamma_k = 1/|K| (Table I).
  • tau_fg, tau_link = 0.5, 0.5
    Inference thresholds for foreground and link binarization (Table I).
  • epsilon = 1e-6
    Numerical stability constant in bottleneck-aware weight normalization (Eq. 12).
assumptions (4)
  • domain assumption Threshold-based connectivity criterion: two points are reachable if there exists a path whose edge weights all exceed a threshold.
    Stated in Section III-A; this is the inference-time criterion the training objective is designed to align with.
  • domain assumption The deterministic skeleton Phi_px faithfully represents the connectivity structure of the union foreground.
    Used in Eq. 2 to construct the support graph V*; skeletonization quality directly affects supervision targets but is not ablated.
  • domain assumption Max-Min subgradient propagation through k-step DP provides a valid and sufficient training signal for connectivity.
    Section III-C relies on almost-everywhere differentiability of min/max; tie-breaking returns one valid subgradient. The paper does not prove convergence properties.
  • domain assumption Bottleneck pixels are the primary cause of topological breaks in slender structure segmentation.
    Stated in Section I and supported by the observation that most topological errors localize in sparse foreground bottlenecks; this motivates the entire approach.
invented entities (3)
  • Topological Gradient Starvation (TGS) independent evidence
    purpose: Names the phenomenon where bottleneck pixels receive insufficient gradients under sum-based losses.
    The paper provides mechanistic evidence via gradient distribution analysis (Fig. 7) showing baseline distributes gradient evenly while +WPRF concentrates it on thin structures. The concept is falsifiable: if gradient distributions were not measurably different, TGS would not hold.
  • Widest-Path Reachability Fields (WPRF) independent evidence
    purpose: The differentiable Max-Min reachability objective and associated training framework.
    Validated on six external datasets with nine architectures; clDice improvements in 47/54 configurations provide independent evidence of effectiveness.
  • OMVIS dataset independent evidence
    purpose: Oral microvessel segmentation dataset introduced for evaluation.
    Publicly available with ethical approval, inter-observer agreement (Dice 0.875), and patient-level splits. Independent of the method itself.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Widest-Path Reachability Fields for Connectivity-Preserving Slender Structure Segmentation." pith.science (2026). https://pith.science/paper/P5E4UPU6

@misc{pith2026260707123,
  author       = {Pith},
  title        = {Pith review of: Widest-Path Reachability Fields for Connectivity-Preserving Slender Structure Segmentation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/P5E4UPU6}},
  note         = {Machine review of arXiv:2607.07123}
}
read the original abstract

Segmenting slender curvilinear structures such as retinal vessels, cracks, and roads demands topological correctness, as even a single-pixel discontinuity can fragment a continuous network and invalidate downstream analysis. Under standard binary-mask supervision, models optimized for pixel-level overlap frequently produce topologically broken predictions. We trace this to a fundamental mismatch: pixel-wise losses distribute gradients uniformly, yet connectivity hinges on a sparse set of bottleneck pixels. These pixels are vastly outnumbered by thick structures and background, rendering their aggregate gradient contribution negligible. We term this phenomenon topological gradient starvation (TGS). To address it, we propose Widest-Path Reachability Fields (WPRF), a differentiable Max-Min reachability objective that redirects gradient flow to connectivity bottlenecks. The module is plug-and-play, backbone-agnostic, and incurs no inference overhead. WPRF implements a differentiable Max-Min objective via dynamic programming on a domain-restricted graph, coupled with a bottleneck-aware observation term that balances gradient contributions across varying structures. Compared to prior topology-aware losses that rely on post-hoc skeletonization or homology computation, WPRF directly optimizes end-to-end reachability via differentiable Max-Min algebra, enabling gradient flow to concentrate on connectivity bottlenecks without auxiliary structures. We introduce OMVIS, a new oral microvessel segmentation dataset. Experiments across nine architectures and six datasets validate the bottleneck-focused gradient routing mechanism. WPRF improves 87\% of experiments with fixed hyperparameters and achieves clDice gains of 7.2 percentage points on structurally fragile datasets.

Figures

Figures reproduced from arXiv: 2607.07123 by the authors.

Figure 1
Figure 1. Qualitative comparison between a standard baseline and its +WPRF [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Architecture of the proposed WPRF framework. A shared backbone predicts union foreground probabilities [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Bottleneck-dominated gradient routing induced by Max-Min reachabil [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (4 more)
Figure 4
Figure 4. Figure 4: Aggregated main results from Tables III–VIII. Per-dataset distribution [PITH_FULL_IMAGE:figures/full_fig_p009_4.png]
Figure 5
Figure 5. Figure 5: Qualitative comparison of nine architectures on DRIVE and Massachusetts Roads datasets. For each sample, top row: baseline prediction; bottom row: [PITH_FULL_IMAGE:figures/full_fig_p010_5.png]
Figure 6
Figure 6. Figure 6: Hyperparameter sensitivity on DRIVE and Massachusetts Roads. Each [PITH_FULL_IMAGE:figures/full_fig_p010_6.png]
Figure 7
Figure 7. Figure 7: Gradient routing diagnosis on a DRIVE training sample. (a) Full [PITH_FULL_IMAGE:figures/full_fig_p011_7.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

76 extracted references · 76 canonical work pages

  1. [1]

    An anatomy- and topology-preserving frame- work for coronary artery segmentation,

    X. Zhang, K. Sun, D. Wu, X. Xiong, J. Liu, L. Yao, S. Li, Y . Wang, J. Feng, and D. Shen, “An anatomy- and topology-preserving frame- work for coronary artery segmentation,”IEEE Transactions on Medical Imaging, vol. 43, no. 2, pp. 723–733, Feb 2024

    work page 2024

  2. [2]

    RSF-Conv: Rotation-and-scale equivariant Fourier parameterized convolution for retinal vessel segmentation,

    Z. Sun, H. Wang, Q. Xie, Y . Zheng, and D. Meng, “RSF-Conv: Rotation-and-scale equivariant Fourier parameterized convolution for retinal vessel segmentation,”IEEE Transactions on Neural Networks and Learning Systems, vol. 36, no. 9, pp. 16 549–16 563, September 2025. [Online]. Available: https://doi.org/10.1109/TNNLS.2025.3560082

    work page doi:10.1109/tnnls.2025.3560082 2025

  3. [3]

    Topology-guided road graph extraction from remote sensing images,

    Y . Zao, Z. Zou, and Z. Shi, “Topology-guided road graph extraction from remote sensing images,”IEEE Transactions on Geoscience and Remote Sensing, vol. 62, pp. 1–14, 2024

    work page 2024

  4. [4]

    Universal vessel segmentation for multi- modality retinal images,

    B. Wen, A. Heinke, A. Agnihotri, D.-U. Bartsch, W. Freeman, T. Nguyen, and C. An, “Universal vessel segmentation for multi- modality retinal images,”IEEE Transactions on Image Processing, vol. 34, pp. 7903–7918, 2025. [Online]. Available: https://doi.org/10. 1109/TIP.2025.3623893

    work page arXiv 2025

  5. [5]

    Aerial images meet crowdsourced trajectories: A new approach to robust road extraction,

    L. Liu, Z. Yang, G. Li, K. Wang, T. Chen, and L. Lin, “Aerial images meet crowdsourced trajectories: A new approach to robust road extraction,”IEEE Transactions on Neural Networks and Learning PREPRINT, JULY 2026 12 Systems, vol. 34, no. 7, pp. 3308–3322, July 2023. [Online]. Available: https://doi.org/10.1109/TNNLS.2022.3141821

    work page doi:10.1109/tnnls.2022.3141821 2026

  6. [6]

    Context enhancing representation for semantic segmentation in remote sensing images,

    L. Fang, P. Zhou, X. Liu, P. Ghamisi, and S. Chen, “Context enhancing representation for semantic segmentation in remote sensing images,” IEEE Transactions on Neural Networks and Learning Systems, vol. 35, no. 3, pp. 4138–4152, March 2024. [Online]. Available: https://doi.org/10.1109/TNNLS.2022.3201820

    work page doi:10.1109/tnnls.2022.3201820 2024

  7. [7]

    Semantic prompt and graph- convolution-structure distillation framework for semantic segmentation of remote sensing images,

    W. Zhou, J. Xie, and C. Xu, “Semantic prompt and graph- convolution-structure distillation framework for semantic segmentation of remote sensing images,”IEEE Transactions on Neural Networks and Learning Systems, pp. 1–14, 2026. [Online]. Available: https: //doi.org/10.1109/TNNLS.2026.3675381

    work page doi:10.1109/tnnls.2026.3675381 2026

  8. [8]

    Sam 2: Segment anything in images and videos,

    N. Ravi, V . Gabeur, Y .-T. Hu, R. Hu, C. Ryali, T. Ma, H. Khedr, R. R ¨adle, C. Rolland, L. Gustafson, E. Mintun, J. Pan, K. V . Alwala, N. Carion, C.-Y . Wu, R. Girshick, P. Dollar, and C. Feichtenhofer, “Sam 2: Segment anything in images and videos,” inInternational Conference on Learning Representations, Y . Yue, A. Garg, N. Peng, F. Sha, and R. Yu, E...

    work page 2025

Show all 76 references
  1. [9]

    Zero-shot capillary segmentation in dermoscopy images via sam2: A case study on oral mucosa,

    W. Su, Y . Zong, R. Jia, J. Qin, and M. Li, “Zero-shot capillary segmentation in dermoscopy images via sam2: A case study on oral mucosa,”IEEE Journal of Biomedical and Health Informatics, pp. 1– 12, 2025

    2025

  2. [10]

    Coarse-to-fine semantic segmentation from image-level labels,

    L. Jing, Y . Chen, and Y . Tian, “Coarse-to-fine semantic segmentation from image-level labels,”IEEE Transactions on Image Processing, vol. 29, pp. 225–236, 2020. [Online]. Available: https://doi.org/10.1109/ TIP.2019.2926748

    2020

  3. [11]

    Spatial structure constraints for weakly supervised semantic segmentation,

    T. Chen, Y . Yao, X. Huang, Z. Li, L. Nie, and J. Tang, “Spatial structure constraints for weakly supervised semantic segmentation,” IEEE Transactions on Image Processing, vol. 33, pp. 1136–1148, 2024. [Online]. Available: https://doi.org/10.1109/TIP.2024.3359041

    2024 doi

  4. [12]

    Guided filter network for semantic image segmentation,

    X. Zhang, W. Zhao, W. Zhang, J. Peng, and J. Fan, “Guided filter network for semantic image segmentation,”IEEE Transactions on Image Processing, vol. 31, pp. 2695–2709, 2022. [Online]. Available: https://doi.org/10.1109/TIP.2022.3160399

    2022 doi

  5. [13]

    Double similarity distillation for semantic image segmentation,

    Y . Feng, X. Sun, W. Diao, J. Li, and X. Gao, “Double similarity distillation for semantic image segmentation,”IEEE Transactions on Image Processing, vol. 30, pp. 5363–5376, 2021. [Online]. Available: https://doi.org/10.1109/TIP.2021.3083113

    2021 doi

  6. [14]

    Geometric boundary guided feature fusion and spatial-semantic context aggregation for semantic segmentation of remote sensing images,

    Y . Wang, H. Zhang, Y . Hu, X. Hu, L. Chen, and S. Hu, “Geometric boundary guided feature fusion and spatial-semantic context aggregation for semantic segmentation of remote sensing images,” IEEE Transactions on Image Processing, vol. 32, pp. 6373–6385, 2023. [Online]. Availab...

    2023 doi

  7. [15]

    Fuzzy attention neural network to tackle discontinuity in airway segmentation,

    Y . Nan, J. Del Ser, Z. Tang, P. Tang, X. Xing, Y . Fang, F. Herrera, W. Pedrycz, S. Walsh, and G. Yang, “Fuzzy attention neural network to tackle discontinuity in airway segmentation,” IEEE Transactions on Neural Networks and Learning Systems, vol. 35, no. 6, pp. 7391–7404, J...

    2024 doi

  8. [16]

    Dynamic snake convolution based on topological geometric constraints for tubular structure segmentation,

    Y . Qi, Y . He, X. Qi, Y . Zhang, and G. Yang, “Dynamic snake convolution based on topological geometric constraints for tubular structure segmentation,” in2023 IEEE/CVF International Conference on Computer Vision (ICCV), October 2023, pp. 6047–6056. [Online]. Available: https...

    2023

  9. [17]

    Skeleton recall loss for connectivity conserving and resource efficient segmentation of thin tubular structures,

    Y . Kirchhoff, M. R. Rokuss, S. Roy, B. Kovacs, C. Ulrich, T. Wald, M. Zenk, P. V ollmuth, J. Kleesiek, F. Isensee, and K. Maier-Hein, “Skeleton recall loss for connectivity conserving and resource efficient segmentation of thin tubular structures,” inComputer Vision – ECCV 20...

    2024 doi

  10. [18]

    Gradient starvation: A learning proclivity in neural networks,

    M. Pezeshki, O. Kaba, Y . Bengio, A. C. Courville, D. Precup, and G. Lajoie, “Gradient starvation: A learning proclivity in neural networks,” inAdvances in Neural Information Processing Systems, M. Ranzato, A. Beygelzimer, Y . Dauphin, P. Liang, and J. W. Vaughan, Eds., vol. 3...

    2021

  11. [19]

    Available: https://proceedings.neurips.cc/paper files/ paper/2021/file/0987b8b338d6c90bbedd8631bc499221-Paper.pdf

    [Online]. Available: https://proceedings.neurips.cc/paper files/ paper/2021/file/0987b8b338d6c90bbedd8631bc499221-Paper.pdf

    2021

  12. [20]

    U-net: Convolutional networks for biomedical image segmentation,

    O. Ronneberger, P. Fischer, and T. Brox, “U-net: Convolutional networks for biomedical image segmentation,” inMedical Image Computing and Computer-Assisted Intervention – MICCAI 2015, N. Navab, J. Horneg- ger, W. M. Wells, and A. F. Frangi, Eds. Cham: Springer International Pu...

    2015

  13. [21]

    FANet: A feedback attention network for improved biomedical image segmentation,

    N. K. Tomar, D. Jha, M. A. Riegler, H. D. Johansen, D. Johansen, J. Rittscher, P. Halvorsen, and S. Ali, “FANet: A feedback attention network for improved biomedical image segmentation,” IEEE Transactions on Neural Networks and Learning Systems, vol. 34, no. 11, pp. 9375–9388,...

    2023 doi

  14. [22]

    SwinPA-Net: Swin transformer-based multiscale feature pyramid aggregation network for medical image segmentation,

    H. Du, J. Wang, M. Liu, Y . Wang, and E. Meijering, “SwinPA-Net: Swin transformer-based multiscale feature pyramid aggregation network for medical image segmentation,”IEEE Transactions on Neural Networks and Learning Systems, vol. 35, no. 4, pp. 5355–5366, April 2024. [Online]...

    2024 doi

  15. [23]

    Equalization loss for long-tailed object recognition,

    J. Tan, C. Wang, B. Li, Q. Li, W. Ouyang, C. Yin, and J. Yan, “Equalization loss for long-tailed object recognition,” in2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), June 2020, pp. 11 659–11 668. [Online]. Available: https: //openaccess.thecvf.com...

    2020

  16. [24]

    Polyloss: A polynomial expansion perspective of classification loss functions,

    Z. Leng, M. Tan, C. Liu, E. D. Cubuk, J. Shi, S. Cheng, and D. Anguelov, “Polyloss: A polynomial expansion perspective of classification loss functions,” inInternational Conference on Learning Representations, 2022. [Online]. Available: https://openreview.net/ forum?id=gSdSJoenupI

    2022

  17. [25]

    Hybrid grid search and bayesian optimization-based random forest regression for predicting material compression pressure in manufacturing processes,

    Y . Zong, Y . Nian, C. Zhang, X. Tang, L. Wang, and L. Zhang, “Hybrid grid search and bayesian optimization-based random forest regression for predicting material compression pressure in manufacturing processes,”Engineering Applications of Artificial Intelligence, vol. 141, p....

    2025

  18. [26]

    Global minimum for active contour models: A minimal path approach,

    L. D. Cohen and R. Kimmel, “Global minimum for active contour models: A minimal path approach,”International journal of computer vision, vol. 24, no. 1, pp. 57–78, aug 1997. [Online]. Available: https://doi.org/10.1023/A:1007922224810

    1997 doi

  19. [27]

    Cape: Connectivity-aware path enforcement loss for curvilinear structure delineation,

    E. Esmaeilzadeh, E. Garaaghaji, F. Hallaji Azad, and D. Oner, “Cape: Connectivity-aware path enforcement loss for curvilinear structure delineation,” inMedical Image Computing and Computer Assisted Intervention – MICCAI 2025, J. C. Gee, D. C. Alexander, J. Hong, J. E. Iglesias...

    2025

  20. [28]

    Available: https://doi.org/10.1007/978-3-032-05162-2 18

    [Online]. Available: https://doi.org/10.1007/978-3-032-05162-2 18

    doi

  21. [29]

    Capturing graphs with hypo-elliptic diffusions,

    C. Toth, D. Lee, C. Hacker, and H. Oberhauser, “Capturing graphs with hypo-elliptic diffusions,” inAdvances in Neural Information Processing Systems, S. Koyejo, S. Mohamed, A. Agarwal, D. Belgrave, K. Cho, and A. Oh, Eds., vol. 35. Curran Associates, Inc., 2022, pp. 38 803–38 ...

    2022

  22. [30]

    Random walks for image segmentation,

    L. Grady, “Random walks for image segmentation,”IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 28, no. 11, pp. 1768– 1783, Nov 2006

    2006

  23. [31]

    Normalized cuts and image segmentation,

    J. Shi and J. Malik, “Normalized cuts and image segmentation,”IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 22, no. 8, pp. 888–905, Aug 2000

    2000

  24. [32]

    Tokencut: Segmenting objects in images and videos with self-supervised transformer and normalized cut,

    Y . Wang, X. Shen, Y . Yuan, Y . Du, M. Li, S. X. Hu, J. L. Crowley, and D. Vaufreydaz, “Tokencut: Segmenting objects in images and videos with self-supervised transformer and normalized cut,”IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 45, no. 12, pp. ...

    2023

  25. [33]

    Cross-dimension affinity distillation for 3d em neuron segmentation,

    X. Liu, M. Cai, Y . Chen, Y . Zhang, T. Shi, R. Zhang, X. Chen, and Z. Xiong, “Cross-dimension affinity distillation for 3d em neuron segmentation,” in2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Los Alamitos, CA, USA: IEEE Computer Society, Jun ...

    2024 doi

  26. [34]

    Efficient neuron segmentation in electron microscopy by affinity-guided queries,

    H. Chen, C. Tang, X. Li, and X. Hu, “Efficient neuron segmentation in electron microscopy by affinity-guided queries,” inThe Thirteenth International Conference on Learning Representations, 2025. [Online]. Available: https://openreview.net/forum?id=Y0QqruhqIa

    2025

  27. [35]

    A meta- contrastive learning hybrid model for adaptive temperature trend prediction in variable ladle preheating,

    Y . Zong, R. Jia, S. Wu, L. Zhang, and D. He, “A meta- contrastive learning hybrid model for adaptive temperature trend prediction in variable ladle preheating,”Engineering Applications of Artificial Intelligence, vol. 162, p. 112750, 2025. [Online]. Available: https://www.sci...

    2025

  28. [36]

    On vanishing gradients, over-smoothing, and over-squashing in GNNs: Bridging recurrent and graph learning,

    ´Alvaro Arroyo, A. Gravina, B. Gutteridge, F. Barbero, C. Gallicchio, X. Dong, M. Bronstein, and P. Vandergheynst, “On vanishing gradients, over-smoothing, and over-squashing in GNNs: Bridging recurrent and graph learning,”arXiv preprint arXiv:2502.10818, 2025. [Online]. Avail...

    2025

  29. [37]

    Sagman: Stability analysis of graph neural networks on the manifolds,

    W. Cheng, C. Deng, A. Aghdaei, Z. Zhang, and Z. Feng, “Sagman: Stability analysis of graph neural networks on the manifolds,” The Thirteenth International Conference on Learning Representations (ICLR), 2024. [Online]. Available: https://openreview.net/forum?id= mVExccNdtK3

    2024

  30. [38]

    Semi-supervised segmentation of histopathology images with noise-aware topological consistency,

    M. Xu, X. Hu, S. Gupta, S. Abousamra, and C. Chen, “Semi-supervised segmentation of histopathology images with noise-aware topological consistency,” inComputer Vision – ECCV 2024, A. Leonardis, E. Ricci, S. Roth, O. Russakovsky, T. Sattler, and G. Varol, Eds. Cham: Springer Na...

    2024 doi

  31. [40]

    Anatomically plausible segmentations: Explicitly preserving topology through prior deformations,

    M. K. Wyburd, N. K. Dinsdale, M. Jenkinson, and A. I. Namburete, “Anatomically plausible segmentations: Explicitly preserving topology through prior deformations,”Medical Image Analysis, vol. 97, p. 103222, July 2024. [Online]. Available: https://www.sciencedirect.com/ science...

    2024

  32. [42]

    Centerline boundary dice loss for vascular segmentation,

    P. Shi, J. Hu, Y . Yang, Z. Gao, W. Liu, and T. Ma, “Centerline boundary dice loss for vascular segmentation,” inMedical Image Computing and Computer Assisted Intervention – MICCAI 2024, M. G. Linguraru, Q. Dou, A. Feragen, S. Giannarou, B. Glocker, K. Lekadir, and J. A. Schna...

    2024 doi

  33. [43]

    Self-supervised 3d skeleton completion for vascular structures,

    J. Ren, Z. Li, W. Cheng, Z. Zou, K. Park, Y . Pan, and H. Ling, “Self-supervised 3d skeleton completion for vascular structures,” inMedical Image Computing and Computer Assisted Intervention – MICCAI 2024, M. G. Linguraru, Q. Dou, A. Feragen, S. Giannarou, B. Glocker, K. Lekad...

    2024 doi

  34. [44]

    Glcp: Global-to-local connectivity preservation for tubular structure segmentation,

    F. Zhou, Z. Gao, H. Zhao, J. Xie, Y . Meng, Y . Zhao, G. Y . H. Lip, and Y . Zheng, “Glcp: Global-to-local connectivity preservation for tubular structure segmentation,” inMedical Image Computing and Computer Assisted Intervention – MICCAI 2025, J. C. Gee, D. C. Alexander, J. ...

    2025 doi

  35. [45]

    Netracer: A topology-aware iterative tracing approach for tubular structure extraction,

    C. Liu, Y . Jiang, and N. Zheng, “Netracer: A topology-aware iterative tracing approach for tubular structure extraction,” inProceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), October 2025, pp. 20 593–20 602

    2025

  36. [46]

    Curvi-tracker: Curvilinear structure segmentation refinement by iterative tracking,

    Z. Heng, M. Pagnucco, E. Meijering, and Y . Song, “Curvi-tracker: Curvilinear structure segmentation refinement by iterative tracking,” Pattern Recognition, vol. 173, p. 112797, May 2026. [Online]. Available: https://www.sciencedirect.com/science/article/pii/S0031320325014608

    2026

  37. [47]

    Boundary-aware axial attention network for high-quality pavement crack detection,

    K. Wu, B. Peng, and D. Zhai, “Boundary-aware axial attention network for high-quality pavement crack detection,”IEEE Transactions on Neural Networks and Learning Systems, vol. 36, no. 7, pp. 13 555–13 566, July 2025. [Online]. Available: https://doi.org/10.1109/ TNNLS.2024.3497145

    2025

  38. [48]

    A deeply supervised convolutional neural network for pavement crack detection with multiscale feature fusion,

    Z. Qu, C. Cao, L. Liu, and D.-Y . Zhou, “A deeply supervised convolutional neural network for pavement crack detection with multiscale feature fusion,”IEEE Transactions on Neural Networks and Learning Systems, vol. 33, no. 9, pp. 4890–4899, September 2022. [Online]. Available:...

    2022 doi

  39. [49]

    Representing topological self-similarity using fractal feature maps for accurate segmentation of tubular structures,

    J. Huang, Y . Zhou, Y . Luo, G. Liu, H. Guo, and G. Yang, “Representing topological self-similarity using fractal feature maps for accurate segmentation of tubular structures,” inComputer Vision – ECCV 2024, A. Leonardis, E. Ricci, S. Roth, O. Russakovsky, T. Sattler, and G. V...

    2024

  40. [50]

    Available: https://doi.org/10.1007/978-3-031-73404-5 9

    [Online]. Available: https://doi.org/10.1007/978-3-031-73404-5 9

    doi

  41. [51]

    Harmonyseg: Tubular structure segmentation with deep-shallow feature fusion and growth-suppression balanced loss,

    Y . Huang, K. Zhang, W. Liu, Y . Wang, V . M. Patel, L. Lu, X. Han, D. Jin, and K. Yan, “Harmonyseg: Tubular structure segmentation with deep-shallow feature fusion and growth-suppression balanced loss,” inProceedings of the IEEE/CVF International Conference on Computer Vision...

    2025

  42. [52]

    Vesselsdf: Distance field priors for vascular network reconstruction,

    S. Esposito, D. Rebain, A. Onken, C. Li, and O. Mac Aodha, “Vesselsdf: Distance field priors for vascular network reconstruction,” inMedical Image Computing and Computer Assisted Intervention – MICCAI 2025, J. C. Gee, D. C. Alexander, J. Hong, J. E. Iglesias, C. H. Sudre, A. V...

    2025 doi

  43. [53]

    A topology-preserving three-stage framework for fully-connected coronary artery extraction,

    Y . Qiu, D. Shan, Y . Wang, P. Dong, D. Wu, X. Yang, Q. Hong, and D. Shen, “A topology-preserving three-stage framework for fully-connected coronary artery extraction,”Medical Image Analysis, vol. 103, p. 103578, July 2025. [Online]. Available: https://www. sciencedirect.com/s...

    2025

  44. [54]

    Topotta: Topology-enhanced test-time adaptation for tubular structure segmentation,

    J. Zhou, W. Wang, S. Li, X. Qu, X. Guo, Y . Liu, W. Tang, X. Lin, and Y . Zheng, “Topotta: Topology-enhanced test-time adaptation for tubular structure segmentation,” inProceedings of the IEEE/CVF International Conference on Computer Vision (ICCV). IEEE, October 2025, pp. 24 1...

    2025

  45. [55]

    cldice - a novel topology- preserving loss function for tubular structure segmentation,

    S. Shit, J. C. Paetzold, A. Sekuboyina, I. Ezhov, A. Unger, A. Zhylka, J. P. W. Pluim, U. Bauer, and B. H. Menze, “cldice - a novel topology- preserving loss function for tubular structure segmentation,” in2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CV...

    2021

  46. [56]

    A topological loss function for deep-learning based image segmentation using persistent homology,

    J. R. Clough, N. Byrne, I. Oksuz, V . A. Zimmer, J. A. Schnabel, and A. P. King, “A topological loss function for deep-learning based image segmentation using persistent homology,”IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 44, no. 12, pp. 8766–8778, dec

  47. [57]

    Available: https://doi.org/10.1109/TPAMI.2020.3013679

    [Online]. Available: https://doi.org/10.1109/TPAMI.2020.3013679

    2020 doi

  48. [58]

    Topology-aware focal loss for 3d image segmentation,

    A. Demir, E. Massaad, and B. Kiziltan, “Topology-aware focal loss for 3d image segmentation,” in2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW). IEEE, June 2023, pp. 580–589. [Online]. Available: https://doi.org/10.1109/ CVPRW59228.2023.00065

    2023

  49. [59]

    Topology-preserving image segmentation with spatial-aware persistent feature matching,

    B. Wen, H. Zhang, D.-U. G. Bartsch, W. Freeman, T. Nguyen, and C. An, “Topology-preserving image segmentation with spatial-aware persistent feature matching,” inProceedings of the IEEE/CVF Inter- national Conference on Computer Vision (ICCV) Workshops, October 2025, pp. 5762–5771

    2025

  50. [60]

    Deep closing: Enhancing topological connectivity in medical tubular segmentation,

    Q. Wu, Y . Chen, W. Liu, X. Yue, and X. Zhuang, “Deep closing: Enhancing topological connectivity in medical tubular segmentation,” IEEE Transactions on Medical Imaging, vol. 43, no. 11, pp. 3990–4003, 2024

    2024

  51. [61]

    Topology optimization in medical image segmentation with fastχeuler characteristic,

    L. Li, Q. Ma, C. Ouyang, J. C. Paetzold, D. Rueckert, and B. Kainz, “Topology optimization in medical image segmentation with fastχeuler characteristic,”IEEE Transactions on Medical Imaging, vol. 44, no. 12, pp. 5221–5232, April 2025

    2025

  52. [62]

    Topology-preserving downsampling of binary images,

    C.-C. Chen and C.-H. Peng, “Topology-preserving downsampling of binary images,” inComputer Vision – ECCV 2024, A. Leonardis, E. Ricci, S. Roth, O. Russakovsky, T. Sattler, and G. Varol, Eds. Cham: Springer Nature Switzerland, 2025, pp. 416–431. [Online]. Available: https://doi...

    2024 doi

  53. [63]

    D. K. Smith,Network Flows: Theory, Algorithms, and Applications. Englewood Cliffs, NJ: Prentice Hall, 1993, vol. 45

    1993

  54. [64]

    Fuzzy connectedness and object definition: Theory, algorithms, and applications in image segmentation,

    J. K. Udupa and S. Samarasekera, “Fuzzy connectedness and object definition: Theory, algorithms, and applications in image segmentation,” Graphical Models and Image Processing, vol. 58, no. 3, pp. 246–261, May 1996. [Online]. Available: https://www.sciencedirect.com/science/ a...

    1996

  55. [65]

    Ridge-based vessel segmentation in color images of the retina,

    J. Staal, M. Abramoff, M. Niemeijer, M. Viergever, and B. van Ginneken, “Ridge-based vessel segmentation in color images of the retina,”IEEE Transactions on Medical Imaging, vol. 23, no. 4, pp. 501–509, 2004. [Online]. Available: https://doi.org/10.1109/TMI.2004. 825627

    2004 doi

  56. [66]

    Octa-500: A retinal dataset for optical coherence tomography angiography study,

    M. Li, K. Huang, Q. Xu, J. Yang, Y . Zhang, Z. Ji, K. Xie, S. Yuan, Q. Liu, and Q. Chen, “Octa-500: A retinal dataset for optical coherence tomography angiography study,”Medical Image Analysis, vol. 93, p. 103092, 2024. [Online]. Available: https://www.sciencedirect.com/scienc...

    2024

  57. [67]

    Deepcrack: Learning hierarchical convolutional features for crack detection,

    Q. Zou, Z. Zhang, Q. Li, X. Qi, Q. Wang, and S. Wang, “Deepcrack: Learning hierarchical convolutional features for crack detection,”IEEE Transactions on Image Processing, vol. 28, no. 3, pp. 1498–1512,

  58. [68]

    Available: https://doi.org/10.1109/TIP.2018.2878966

    [Online]. Available: https://doi.org/10.1109/TIP.2018.2878966

    2018 doi

  59. [69]

    Learning to detect roads in high- resolution aerial images,

    V . Mnih and G. E. Hinton, “Learning to detect roads in high- resolution aerial images,” inComputer Vision – ECCV 2010, K. Daniilidis, P. Maragos, and N. Paragios, Eds. Berlin, Heidelberg: Springer Berlin Heidelberg, 2010, pp. 210–223. [Online]. Available: https://doi.org/10.1...

    2010 doi

  60. [70]

    Encoder-decoder with atrous separable convolution for semantic image segmentation,

    L.-C. Chen, Y . Zhu, G. Papandreou, F. Schroff, and H. Adam, “Encoder-decoder with atrous separable convolution for semantic image segmentation,” inComputer Vision – ECCV 2018, V . Ferrari, M. Hebert, C. Sminchisescu, and Y . Weiss, Eds. Cham: Springer International Publishing...

    2018 doi

  61. [71]

    Segformer: Simple and efficient design for semantic segmentation with transformers,

    E. Xie, W. Wang, Z. Yu, A. Anandkumar, J. M. Alvarez, and P. Luo, “Segformer: Simple and efficient design for semantic segmentation with transformers,” inAdvances in Neural Information Processing Systems, M. Ranzato, A. Beygelzimer, Y . Dauphin, P. Liang, and J. W. Vaughan, Ed...

    2021

  62. [72]

    Masked-attention mask transformer for universal image segmentation,

    B. Cheng, I. Misra, A. G. Schwing, A. Kirillov, and R. Girdhar, “Masked-attention mask transformer for universal image segmentation,” in2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). IEEE, June 2022, pp. 1280–1289. [Online]. Available: https://doi....

    2022 doi

  63. [73]

    Transunet: Rethinking the u-net architecture design for medical image segmentation through the lens of transformers,

    J. Chen, J. Mei, X. Li, Y . Lu, Q. Yu, Q. Wei, X. Luo, Y . Xie, E. Adeli, Y . Wang, M. P. Lungren, S. Zhang, L. Xing, L. Lu, A. Yuille, and Y . Zhou, “Transunet: Rethinking the u-net architecture design for medical image segmentation through the lens of transformers,” Medical ...

    2024

  64. [74]

    nnu-net: a self-configuring method for deep learning-based biomedical image segmentation,

    F. Isensee, P. F. Jaeger, S. A. A. Kohl, J. Petersen, and K. H. Maier-Hein, “nnu-net: a self-configuring method for deep learning-based biomedical image segmentation,”Nature Methods, vol. 18, no. 2, pp. 203–211,

  65. [75]

    Available: https://doi.org/10.1038/s41592-020-01008-z

    [Online]. Available: https://doi.org/10.1038/s41592-020-01008-z

    doi

  66. [76]

    Cs-net: Channel and spatial attention network for curvilinear structure segmentation,

    L. Mou, Y . Zhao, L. Chen, J. Cheng, Z. Gu, H. Hao, H. Qi, Y . Zheng, A. Frangi, and J. Liu, “Cs-net: Channel and spatial attention network for curvilinear structure segmentation,” inMedical Image Computing and Computer Assisted Intervention – MICCAI 2019, D. Shen, T. Liu, T. ...

    2019 doi

  67. [77]

    Swin-unet: Unet-like pure transformer for medical image segmentation,

    H. Cao, Y . Wang, J. Chen, D. Jiang, X. Zhang, Q. Tian, and M. Wang, “Swin-unet: Unet-like pure transformer for medical image segmentation,” inComputer Vision – ECCV 2022 Workshops, L. Karlinsky, T. Michaeli, and K. Nishino, Eds. Cham: Springer Nature Switzerland, 2023, pp. 20...

    2022 doi

  68. [78]

    U-mamba: Enhancing long-range dependency for biomedical image segmentation,

    J. Ma, F. Li, and B. Wang, “U-mamba: Enhancing long-range dependency for biomedical image segmentation,”arXiv preprint arXiv:2401.04722, 2024. [Online]. Available: https://arxiv.org/abs/2401. 04722

    2024 arXiv

Pith tools

Reviewed July 9, 2026 · model on record in the stance chip above.