Pith. sign in

REVIEW 3 major objections 5 minor 69 references

Retinal Vessel Segmentation via Neuron Programming

T0 review · 3 major / 5 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read This paper claims that searching the neuron type at each network position jointly with the architecture, not just the architecture alone, yields better retinal vessel segmentation, with a 0.84M-parameter searched network achieving the…

desk verdict Solid empirical NAS paper whose central neuron-diversity claim rests on single-run ablations; send it to review, but ask for error bars and code. read the letter →

arxiv 2411.11110 v1 pith:TDQNWH46 submitted 2024-11-17 eess.IV cs.CV

Tingting Wu , Ruyi Min , Peixuan Song , Hengtao Guo , Tieyong Zeng , Feng-Lei Fan This is my paper · ORCID
classification eess.IVcs.CV
keywords retinalvesselsegmentationneuronprogrammingquadraticneuralarchitecturesearchhypernetworkgeneticalgorithmneuronaldiversityU-Net
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper proposes neuron programming: instead of fixing every neuron in a network to one type, it searches over neuron types (conventional inner-product neurons versus quadratic neurons) jointly with the architecture, using a genetic algorithm on a U-Net backbone. On DRIVE the resulting 0.84M-parameter network achieves the highest sensitivity (0.8489) and F1 (0.8334) among the compared methods, including much larger models, and transfers to STARE, CHASE DB1, and HRF with competitive F1. The authors argue that this shows no single universal neuron type suits all locations, and that treating neuron type as a searchable gene complements architecture-level NAS. They also build a hypernetwork trained on the best searched configurations to predict neuron types for unseen depth and channel settings, cutting search cost.

What carries the argument

The carrying mechanism is the genetic-algorithm search over a block-based encoding in which each node's operation gene includes kernel size, normalization, activation order, and neuron type (conventional or quadratic), giving 16 operation sequences per node. Quadratic neurons replace the inner product with a product of two inner products plus a Hadamard-square term, following Fan et al. (2018), and are trained with the ReLinear initialization of Fan et al. (2023) to keep higher-order terms stable. The hypernetwork is a small classifier taking network depth and channel size as input and predicting the operation sequence at each position, trained with categorical cross-entropy on the top-10 searched configurations per depth and channel group. What makes the argument run is that the neuron-type gene can be flipped while everything else stays fixed, letting the ablations in Table 8 isolate the contribution of neuronal diversity.

What would settle it

Repeat the three neuron-type configurations from Table 8 on DRIVE across at least 20 random seeds and test whether the mixed network's F1 and AUC are consistently above both single-type networks with a paired test; if not, the neuron-type gene is not the source of the reported gain. A second check: evaluate hypernetwork predictions at held-out channels 8, 18, and 28 and report F1 with variance against the top-10 searched networks at those channels.

Watch

Extended reading notes

Core claim

The paper's central claim is that searching the neuron type of each position alongside the network architecture—rather than training a homogeneous network of one neuron kind—improves segmentation performance, and that this 'neuron programming' is complementary to architecture search. Concretely, the best joint search returns a U-shaped network that mixes conventional and quadratic neurons, where the quadratic neuron computes $y = \sigma((w_1^\top x + b_1)(w_2^\top x + b_2) + w_3^\top (x \odot x) + b_3)$, and this mixture reaches DRIVE SE 0.8489, F1 0.8334, and AUC 0.9901, the best sensitivity and F1 among all compared models while using only 0.84M parameters. The same searched network transfers to STARE (F1 0.8692), CHASE DB1 (F1 0.8224), and HRF (F1 0.8123). The paper further claims that the hypernetwork-predicted networks perform comparably to NAS-found optima and better than the original U-Net, supporting the view that neuron-type assignment can be predicted rather than searched at deployment time.

Load-bearing premise

The paper treats the F1 differences in Table 8—0.8334 for the mixed network versus 0.8265 and 0.8300 for single-type networks—as evidence that neuron diversity itself carries the gain, but each number comes from a single run, so the load-bearing assumption is that those gaps exceed run-to-run noise.

Editorial extensions

If this is right

  • A single searched network of 0.84M parameters outperforms or matches models with 5 to 43M parameters on four benchmark sets, so neuron-type search is a parameter-efficient route to high sensitivity on fine vessels.
  • Because the plug-and-play variant first searches architecture then neuron types, any existing well-performing architecture can be upgraded to heterogeneous neurons without re-running a full NAS.
  • The hypernetwork turns neuron programming into a one-shot prediction: given depth and channel settings, it outputs the whole configuration, eliminating search at deployment time.
  • The combination of focal loss and a searched mixture of neuron types keeps sensitivity high on tiny vessels, the clinically relevant failure mode for diabetic retinopathy screening.

Reading between the lines

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

  • An untested consequence is that neuron-type search may transfer to other dense prediction tasks, such as vessel segmentation in OCT angiography or coronary angiography, whenever the searched network is retrained from scratch, since the claim is about the neuron-type gene rather than DRIVE-specific weights.
  • The paper's frequency analysis suggests quadratic neurons are especially frequent in later blocks; a testable extension is to restrict the search space to later blocks only and measure whether most of the F1 gain remains.
  • The hypernetwork's interpolation claim is supported only at three held-out points without reported numbers; a stronger validation would compare predicted configurations against the full search distribution at those channels, with standard deviations.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 5 minor

Summary. The paper introduces "neuron programming": treating neuron type (conventional versus quadratic) as an additional searchable design dimension, complementary to architecture-level NAS. A genetic algorithm jointly searches architecture and neuron-type assignments within a U-shaped encoder–decoder on the DRIVE training set, yielding a 0.84M-parameter network that reports state-of-the-art DRIVE SE/F1/AUC and transfers to STARE, CHASE DB1, and HRF. A plug-and-play variant first searches the architecture and then optimizes neuron types, and a hypernetwork is trained to predict neuron configurations for unseen depth/channel settings. The paper claims that mixing neuron types outperforms single-neuron-type networks and that the hypernetwork predictions are comparable to full NAS results at a fraction of the cost.

Significance. If the claims hold, the paper adds a genuinely new and cheap design dimension to neural architecture search for medical image segmentation, with strong empirical results on four public benchmarks and a parameter count an order of magnitude below competing methods. The search is run only on DRIVE and the found network is tested on externally fixed splits of STARE, CHASE DB1, and HRF, which is a commendable evaluation design. The hypernetwork is evaluated on depth/channel combinations (8, 18, 28) that lie inside the range but off the 5-step training grid of channel sizes, so the generalization test is a genuine holdout rather than a re-reporting of training points. The central limiting factor is that the key isolation of the neuron-mixing effect (Table 8) rests on single runs with no variance reporting, and the hypernetwork comparison is reported only graphically.

major comments (3)
  1. [§5, Table 8] The paper's central claim that neuronal diversity, rather than the architecture search itself, drives the gains rests on Table 8, where replacing the searched mixture with homogeneous conventional or quadratic neurons lowers F1 from 0.8334 to 0.8265 and 0.8300 and AUC from 0.9901 to 0.9879 and 0.9892. These margins (0.0034–0.0069 in F1) are each based on a single training run with no standard deviation, no seed count, and no significance test. Because DRIVE's test set has only 20 images and vessel-segmentation results are known to vary with initialization at this level, the reported differences are within plausible run-to-run noise; the paper needs repeated-seed results (means and variances, or paired tests) to support the assertion that the neuron-type gene is not decorative.
  2. [§4.4, Figure 13] The claim that "the hypernetwork-predicted network performs comparably to optimal networks derived through NAS and outperforms the original U-Net" is supported only by three bar charts (depth/channel = 2/8, 3/18, 4/28) with no numeric table and no error bars. The reader cannot verify the magnitude of the claimed parity or the improvement over U-Net, nor can the holdout nature of channels 8, 18, and 28 be quantitatively assessed. A table reporting ACC/SE/SP/F1/AUC for the hypernetwork prediction, the NAS-searched network, and the U-Net baseline at these three settings should be added.
  3. [§4.2, Tables 4–7] The comparisons against prior state-of-the-art methods are reported as single point estimates for all methods, including the proposed one, with no variance or significance information. This is less serious than the Table 8 issue because the margins to the strongest competitors (e.g., DRIVE SE 0.8489 vs 0.8356 for FR-UNet; F1 0.8334 vs 0.8316) are again small and could overlap under seed variation. The authors should state whether the reported numbers for the proposed method are means over multiple runs and, at minimum, provide the number of seeds used for the main results.
minor comments (5)
  1. [§4.1, Eq. (2)] In the focal loss equation, the exponent is written as omega (ω) but the prose two sentences later refers to it as w; the notation should be unified.
  2. [§3.3] The hypernetwork is trained on "the top 10 architectures with the highest F1 scores" per depth/channel group, but the total number of architectures per group, the F1 range of the selected subset, and the split of the resulting training data are not given; this information is needed to judge whether the hypernetwork is learning a meaningful mapping or memorizing a small sample.
  3. [§4.2, Figure 7] The frequency counts in Figure 7 are presented without the total number of evaluated architectures per depth, so the reader cannot tell whether a frequency of 271 out of 1000 means a robust preference or a minority pattern; please add denominators.
  4. [§4.1] The genetic algorithm parameters (population size 20, 50 generations, crossover 0.9, mutation 0.7, bit-flip 0.5) are given without any sensitivity or repeat-run analysis; since these are free parameters of the search, at least a brief discussion of their influence or a citation to prior GA-NAS usage would strengthen the reproducibility story.
  5. [Throughout] The paper contains many minor grammatical errors and awkward formulations (e.g., "in a dedicate manner," "towards a condensed neuron programming," "the hypernetwork 'programs' in real-time and convenient manner"). These do not affect the technical content but should be corrected in a revision.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the central results are externally benchmarked and the hypernetwork predictions are genuine holdouts.

full rationale

The claimed derivation chain is empirical rather than deductive. The reported DRIVE, STARE, CHASE DB1, and HRF results are evaluated on fixed public test splits after architecture search on DRIVE training data, so the test metrics are not the optimized objective. The hypernetwork is trained on top-10 joint-search configurations for depths 2-4 and channel widths 5, 10, 15, 20, 25, 30, 35, and then tested at channel widths 8, 18, and 28; these widths are absent from the training grid, making the predicted configurations genuine holdouts rather than fitted inputs. Table 8's diversity ablation compares a searched heterogeneous network against two unsearched homogeneous re-trainings; the margin is not guaranteed by construction because the reported F1 and AUC are test-set values, not the search fitness, so any weakness there is statistical (single run, no error bars), not circular. The self-citations to Fan et al. and Liao et al. motivate the choice of quadratic neurons and heterogeneous expressivity, but the cited theorem is a parameter-free existence result about a function class, not an assumption that entails the segmentation outcomes, and the empirical benchmarks do not reduce to it. No equation in the paper is shown to equal its own input, no fitted parameter is renamed as a prediction, and no uniqueness theorem is invoked to rule out alternatives.

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

The method rests on hand-chosen search and training hyperparameters (GA population, crossover and mutation rates, top-10 selection, ReLinear learning-rate split), an unreported focal-loss configuration, and the choice of quadratic neuron taken from the authors' prior work. No new mathematical axioms are introduced: the heterogeneity expressiveness result is cited from Liao et al. (2024). The main domain assumptions are that training-data fitness transfers to unseen test sets and that the (depth, channel) to configuration mapping learned by the hypernetwork interpolates to held-out channels.

free parameters (5)
  • GA population size and generation count = N=20, T=50
    Hand-chosen in Section 4.1; directly controls how many architectures are scored and hence which designs the hypernetwork later sees.
  • GA crossover, mutation, and bit-flip probabilities = 0.9, 0.7, 0.5
    Hand-chosen in Section 4.1; these probabilities shape the search trajectory and the diversity of the top-10 training set.
  • Focal loss balance factors = not reported
    Eq. 2 defines alpha and omega, but Section 4.1 never gives their values; every quantitative result depends on them.
  • Top-k architectures per parameter group for hypernetwork training = k=10
    Hand-chosen threshold in Section 3.3; changing k changes the hypernetwork's training distribution and its notion of a good configuration.
  • ReLinear learning rate split and Lookahead parameters = 0.001 vs 0.0001; alpha 0.05, k 6
    Adopted from Fan et al. (2023) in Section 4.1 to stabilize quadratic networks; the comparison assumes this scheme is neutral to both neuron types.
assumptions (5)
  • domain assumption The quadratic neuron of Eq. 1 is the right alternative neuron type; O(n^2) quadratic designs are excluded for scalability (Section 2.2).
    Modeling choice inherited from the authors' prior work (Fan et al. 2018); the paper argues rather than proves this neuron should be the search dimension.
  • standard math Heterogeneous conventional/quadratic networks can approximate some functions with polynomially fewer neurons than homogeneous networks (Liao et al. 2024, cited in Section 1).
    Cited theorem used as motivation; the paper does not verify it applies to the searched U-Nets on vessel data.
  • domain assumption Fitness computed on DRIVE training data predicts test performance and transfers to STARE, CHASE DB1, and HRF (Sections 3.1, 4.1).
    All reported gains rest on this transfer; no validation split during search is described, so the selection step may overfit the training set.
  • domain assumption ReLU plus ReLinear initialization keeps quadratic networks trainable (Section 4.1, citing Fan et al. 2023).
    Training stability is assumed from prior work; the paper does not ablate this initialization choice.
  • ad hoc to paper Instance normalization aids quadratic neurons because it curbs feature magnitude, given the degree explosion of deep quadratic networks (Section 4.2).
    Post-hoc explanation for why the search favored instance normalization; not tested directly.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Retinal Vessel Segmentation via Neuron Programming." pith.science (2026). https://pith.science/paper/TDQNWH46

@misc{pith2026241111110,
  author       = {Pith},
  title        = {Pith review of: Retinal Vessel Segmentation via Neuron Programming},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/TDQNWH46}},
  note         = {Machine review of arXiv:2411.11110}
}
read the original abstract

The accurate segmentation of retinal blood vessels plays a crucial role in the early diagnosis and treatment of various ophthalmic diseases. Designing a network model for this task requires meticulous tuning and extensive experimentation to handle the tiny and intertwined morphology of retinal blood vessels. To tackle this challenge, Neural Architecture Search (NAS) methods are developed to fully explore the space of potential network architectures and go after the most powerful one. Inspired by neuronal diversity which is the biological foundation of all kinds of intelligent behaviors in our brain, this paper introduces a novel and foundational approach to neural network design, termed ``neuron programming'', to automatically search neuronal types into a network to enhance a network's representation ability at the neuronal level, which is complementary to architecture-level enhancement done by NAS. Additionally, to mitigate the time and computational intensity of neuron programming, we develop a hypernetwork that leverages the search-derived architectural information to predict optimal neuronal configurations. Comprehensive experiments validate that neuron programming can achieve competitive performance in retinal blood segmentation, demonstrating the strong potential of neuronal diversity in medical image analysis.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

69 extracted references · 64 canonical work pages

  1. [1]

    Z., Yakopcic, C., Hasan, M., Taha, T

    Alom, M. Z., Yakopcic, C., Hasan, M., Taha, T. M., & Asari, V. K. (2019). Recurrent residual u-net for medical image segmentation. Journal of medical imaging , 6(1), 014006--014006

    work page 2019

  2. [2]

    & Trucco, E

    Annunziata, R. & Trucco, E. (2016). Accelerating convolutional sparse coding for curvilinear structures segmentation by refining scird-ts filter banks. IEEE transactions on medical imaging , 35(11), 2381--2392

    work page 2016

  3. [3]

    M., & Weston, N

    Brock, A., Lim, T., Ritchie, J. M., & Weston, N. (2017). Smash: one-shot model architecture search through hypernetworks. arXiv preprint arXiv:1708.05344

    arXiv 2017

  4. [4]

    & Karpatne, A

    Bu, J. & Karpatne, A. (2021). Quadratic residual networks: A new class of neural networks for solving forward and inverse problems in physics involving pdes. In Proceedings of the 2021 SIAM International Conference on Data Mining (SDM) (pp.\ 675--683).: SIAM

    work page 2021

  5. [5]

    K., Zhou, J., Lu, P., Molaei, S., & Clifton, D

    Chauhan, V. K., Zhou, J., Lu, P., Molaei, S., & Clifton, D. A. (2024). A brief review of hypernetworks in deep learning. Artificial Intelligence Review , 57(9), 1--29

    work page 2024

  6. [6]

    G., Moschoglou, S., Bouritsas, G., Deng, J., Panagakis, Y., & Zafeiriou, S

    Chrysos, G. G., Moschoglou, S., Bouritsas, G., Deng, J., Panagakis, Y., & Zafeiriou, S. (2021). Deep polynomial neural networks. IEEE transactions on pattern analysis and machine intelligence , 44(8), 4021--4034

    work page 2021

  7. [7]

    H., & Hutter, F

    Elsken, T., Metzen, J. H., & Hutter, F. (2019). Neural architecture search: A survey. The Journal of Machine Learning Research , 20(1), 1997--2017

    work page 2019

  8. [8]

    Fan, F., Cong, W., & Wang, G. (2018). A new type of neurons for machine learning. International journal for numerical methods in biomedical engineering , 34(2), e2920

    work page 2018

Show all 69 references
  1. [9]

    K., Singh, R., Qian, G., Getzin, M., Teng, Y., Hahn, J., & Wang, G

    Fan, F., Shan, H., Kalra, M. K., Singh, R., Qian, G., Getzin, M., Teng, Y., Hahn, J., & Wang, G. (2019). Quadratic autoencoder (q-ae) for low-dose ct denoising. IEEE transactions on medical imaging , 39(6), 2035--2050

    2019

  2. [10]

    Fan, F.-L., Li, M., Wang, F., Lai, R., & Wang, G. (2023). On expressivity and trainability of quadratic networks. IEEE Transactions on Neural Networks and Learning Systems

    2023

  3. [11]

    M., Remagnino, P., Hoppe, A., Uyyanonvara, B., Rudnicka, A

    Fraz, M. M., Remagnino, P., Hoppe, A., Uyyanonvara, B., Rudnicka, A. R., Owen, C. G., & Barman, S. A. (2012). Blood vessel segmentation methodologies in retinal images--a survey. Computer methods and programs in biomedicine , 108(1), 407--433

    2012

  4. [12]

    Giarratano, Y., Bianchi, E., Gray, C., Morris, A., MacGillivray, T., Dhillon, B., & Bernabeu, M. O. (2020). Automated segmentation of optical coherence tomography angiography images: benchmark data and clinically relevant metrics. Translational vision science & technology , 9(...

    2020

  5. [13]

    Goyal, M., Goyal, R., & Lall, B. (2020). Improved polynomial neural networks with normalised activations. In 2020 International Joint Conference on Neural Networks (IJCNN) (pp.\ 1--8).: IEEE

    2020

  6. [14]

    P., & Ahmad, S

    Grewal, K., Forest, J., Cohen, B. P., & Ahmad, S. (2021). Going beyond the point neuron: active dendrites and sparse representations for continual learning. bioRxiv , (pp.\ 2021--10)

    2021

  7. [15]

    Guo, C., Szemenyei, M., Yi, Y., Wang, W., Chen, B., & Fan, C. (2021). Sa-unet: Spatial attention u-net for retinal vessel segmentation. In 2020 25th international conference on pattern recognition (ICPR) (pp.\ 1236--1242).: IEEE

    2021

  8. [16]

    & Kembhavi, A

    Gupta, T. & Kembhavi, A. (2023). Visual programming: Compositional visual reasoning without training. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (pp.\ 14953--14962)

    2023

  9. [17]

    Ha, D., Dai, A., & Le, Q. V. (2016). Hypernetworks. arXiv preprint arXiv:1609.09106

    2016 arXiv

  10. [18]

    Hoover, A., Kouznetsova, V., & Goldbaum, M. (2000). Locating blood vessels in retinal images by piecewise threshold probing of a matched filter response. IEEE Transactions on Medical imaging , 19(3), 203--210

    2000

  11. [19]

    M., & Ryan, C

    Houreh, Y., Mahdinejad, M., Naredo, E., Dias, D. M., & Ryan, C. (2021). Hnas: Hyper neural architecture search for image segmentation. In ICAART (2) (pp.\ 246--256)

    2021

  12. [20]

    Izhikevich, E. M. (2003). Simple model of spiking neurons. IEEE Transactions on neural networks , 14(6), 1569--1572

    2003

  13. [21]

    P., Behabadi, B

    Jadi, M. P., Behabadi, B. F., Poleg-Polsky, A., Schiller, J., & Mel, B. W. (2014). An augmented two-layer model captures nonlinear analog spatial integration effects in pyramidal neuron dendrites. Proceedings of the IEEE , 102(5), 782--798

    2014

  14. [22]

    & Mojon, D

    Jiang, X. & Mojon, D. (2003). Adaptive local thresholding by verification-based multithreshold probing with application to vessel detection in retinal images. IEEE Transactions on Pattern Analysis and Machine Intelligence , 25(1), 131--137

    2003

  15. [23]

    Jiang, Y., Yang, F., Zhu, H., Zhou, D., & Zeng, X. (2020). Nonlinear cnn: improving cnns with quadratic convolutions. Neural Computing and Applications , 32(12), 8507--8516

    2020

  16. [24]

    D., Chen, Q., Wei, L., & Su, R

    Jin, Q., Meng, Z., Pham, T. D., Chen, Q., Wei, L., & Su, R. (2019). Dunet: A deformable network for retinal vessel segmentation. Knowledge-Based Systems , 178, 149--162

    2019

  17. [25]

    & Gooya, A

    Kalaie, S. & Gooya, A. (2017). Vascular tree tracking and bifurcation points detection in retinal images using a hierarchical probabilistic model. Computer methods and programs in biomedicine , 151, 139--149

    2017

  18. [26]

    B., Siddique, M

    Khan, K. B., Siddique, M. S., Ahmad, M., & Mazzara, M. (2020). A hybrid unsupervised approach for retinal vessel segmentation. BioMed Research International , 2020(1), 8365783

    2020

  19. [27]

    Kingma, D. P. & Ba, J. (2014). Adam: A method for stochastic optimization. arXiv preprint arXiv:1412.6980

    2014 arXiv

  20. [28]

    F., Odstr c ilik, J., Michelson, G., & Hornegger, J

    K \"o hler, T., Budai, A., Kraus, M. F., Odstr c ilik, J., Michelson, G., & Hornegger, J. (2013). Automatic no-reference quality assessment for retinal fundus images using vessel segmentation. In Proceedings of the 26th IEEE international symposium on computer-based medical sy...

    2013

  21. [29]

    & Kiraz, B

    Ku s , Z. & Kiraz, B. (2023). Evolutionary architecture optimization for retinal vessel segmentation. IEEE Journal of Biomedical and Health Informatics

    2023

  22. [30]

    Le, T., Nguyen, D., & Le, B. (2023). Generating relation-specific weights for convkb using a hypernetwork architecture. Applied Intelligence , 53(18), 21092--21115

    2023

  23. [31]

    & Rahardja, S

    Li, D. & Rahardja, S. (2021). Bseresu-net: An attention-based before-activation residual u-net for retinal vessel segmentation. Computer Methods and Programs in Biomedicine , 205, 106070

    2021

  24. [32]

    Li, K., Qi, X., Luo, Y., Yao, Z., Zhou, X., & Sun, M. (2020). Accurate retinal vessel segmentation in color fundus images via fully attention-based networks. IEEE Journal of Biomedical and Health Informatics , 25(6), 2071--2081

    2020

  25. [33]

    Li, Q., Feng, B., Xie, L., Liang, P., Zhang, H., & Wang, T. (2015). A cross-modality learning approach for vessel segmentation in retinal images. IEEE transactions on medical imaging , 35(1), 109--118

    2015

  26. [34]

    Li, Y., Zhang, Y., Liu, J.-Y., Wang, K., Zhang, K., Zhang, G.-S., Liao, X.-F., & Yang, G. (2022). Global transformer and dual local attention network via deep-shallow hierarchical feature fusion for retinal vessel segmentation. IEEE Transactions on Cybernetics

    2022

  27. [35]

    Liao, J.-X., Hou, B.-J., Dong, H.-C., Zhang, H., Zhang, X., Sun, J., Zhang, S., & Fan, F.-L. (2024). Quadratic neuron-empowered heterogeneous autoencoder for unsupervised anomaly detection. IEEE Transactions on Artificial Intelligence

    2024

  28. [36]

    Lin, T.-Y., Goyal, P., Girshick, R., He, K., & Doll \'a r, P. (2017). Focal loss for dense object detection. In Proceedings of the IEEE international conference on computer vision (pp.\ 2980--2988)

    2017

  29. [37]

    Liu, W., Yang, H., Tian, T., Cao, Z., Pan, X., Xu, W., Jin, Y., & Gao, F. (2022). Full-resolution network and dual-threshold iteration for retinal vessel and coronary angiograph segmentation. IEEE journal of biomedical and health informatics , 26(9), 4623--4634

    2022

  30. [38]

    & Shah, S

    Mantini, P. & Shah, S. K. (2021). Cqnn: Convolutional quadratic neural networks. In 2020 25th International Conference on Pattern Recognition (ICPR) (pp.\ 9819--9826).: IEEE

    2021

  31. [39]

    Micikevicius, P., Narang, S., Alben, J., Diamos, G., Elsen, E., Garcia, D., Ginsburg, B., Houston, M., Kuchaiev, O., Venkatesh, G., et al. (2018). Mixed precision training. In International Conference on Learning Representations

    2018

  32. [40]

    Mou, L., Zhao, Y., Chen, L., Cheng, J., Gu, Z., Hao, H., Qi, H., Zheng, Y., Frangi, A., & Liu, J. (2019). Cs-net: Channel and spatial attention network for curvilinear structure segmentation. In Medical Image Computing and Computer Assisted Intervention--MICCAI 2019: 22nd Inte...

    2019

  33. [41]

    F., et al

    Mou, L., Zhao, Y., Fu, H., Liu, Y., Cheng, J., Zheng, Y., Su, P., Yang, J., Chen, L., Frangi, A. F., et al. (2021). Cs2-net: Deep learning segmentation of curvilinear structures in medical imaging. Medical image analysis , 67, 101874

    2021

  34. [42]

    Peng, H., Xie, P., Liu, L., Kuang, X., Wang, Y., Qu, L., Gong, H., Jiang, S., Li, A., Ruan, Z., et al. (2021). Morphological diversity of single neurons in molecularly defined cell types. Nature , 598(7879), 174--181

    2021

  35. [43]

    D., Naredo, E., & Ryan, C

    Popat, V., Mahdinejad, M., Cede \ n o, O. D., Naredo, E., & Ryan, C. (2020). Ga-based u-net architecture optimization applied to retina blood vessel segmentation. In IJCCI (pp.\ 192--199)

    2020

  36. [44]

    Rajesh, C., Sadam, R., & Kumar, S. (2023). An evolutionary u-shaped network for retinal vessel segmentation using binary teaching--learning-based optimization. Biomedical Signal Processing and Control , 83, 104669

    2023

  37. [45]

    & Perfetti, R

    Ricci, E. & Perfetti, R. (2007). Retinal blood vessel segmentation using line operators and support vector classification. IEEE transactions on medical imaging , 26(10), 1357--1365

    2007

  38. [46]

    Ronneberger, O., Fischer, P., & Brox, T. (2015). U-net: Convolutional networks for biomedical image segmentation. In Medical Image Computing and Computer-Assisted Intervention--MICCAI 2015: 18th International Conference, Munich, Germany, October 5-9, 2015, Proceedings, Part II...

    2015

  39. [47]

    Shen, N., Xu, T., Bian, Z., Huang, S., Mu, F., Huang, B., Xiao, Y., & Li, J. (2022). Scanet: A unified semi-supervised learning framework for vessel segmentation. IEEE Transactions on Medical Imaging

    2022

  40. [48]

    Shen, N., Xu, T., Huang, S., Mu, F., & Li, J. (2023). Expert-guided knowledge distillation for semi-supervised vessel segmentation. IEEE Journal of Biomedical and Health Informatics

    2023

  41. [49]

    V., Leandro, J

    Soares, J. V., Leandro, J. J., Cesar, R. M., Jelinek, H. F., & Cree, M. J. (2006). Retinal vessel segmentation using the 2-d gabor wavelet and supervised classification. IEEE Transactions on medical Imaging , 25(9), 1214--1222

    2006

  42. [50]

    D., Niemeijer, M., Viergever, M

    Staal, J., Abr \`a moff, M. D., Niemeijer, M., Viergever, M. A., & Van Ginneken, B. (2004). Ridge-based vessel segmentation in color images of the retina. IEEE transactions on medical imaging , 23(4), 501--509

    2004

  43. [51]

    Sun, J.-D., Yao, C., Liu, J., Liu, W., & Yu, Z.-K. (2022). Gnas-u 2 net: a new optic cup and optic disc segmentation architecture with genetic neural architecture search. IEEE Signal Processing Letters , 29, 697--701

    2022

  44. [52]

    Tsapanos, N., Tefas, A., Nikolaidis, N., & Pitas, I. (2018). Neurons with paraboloid decision boundaries for improved neural network classification performance. IEEE transactions on neural networks and learning systems , 30(1), 284--294

    2018

  45. [53]

    Wang, W., Zhong, J., Wu, H., Wen, Z., & Qin, J. (2020). Rvseg-net: An efficient feature pyramid cascade network for retinal vessel segmentation. In Medical Image Computing and Computer Assisted Intervention--MICCAI 2020: 23rd International Conference, Lima, Peru, October 4--8,...

    2020

  46. [54]

    Wei, J., Zhu, G., Fan, Z., Liu, J., Rong, Y., Mo, J., Li, W., & Chen, X. (2021). Genetic u-net: automatically designed deep networks for retinal vessel segmentation using a genetic algorithm. IEEE Transactions on Medical Imaging , 41(2), 292--307

    2021

  47. [55]

    Weng, Y., Zhou, T., Li, Y., & Qiu, X. (2019). Nas-unet: Neural architecture search for medical image segmentation. IEEE access , 7, 44247--44257

    2019

  48. [56]

    Wu, H., Wang, W., Zhong, J., Lei, B., Wen, Z., & Qin, J. (2021). Scs-net: A scale and context sensitive network for retinal vessel segmentation. Medical Image Analysis , 70, 102025

    2021

  49. [57]

    Wu, Y., Xia, Y., Song, Y., Zhang, D., Liu, D., Zhang, C., & Cai, W. (2019). Vessel-net: Retinal vessel segmentation under multi-path supervision. In Medical Image Computing and Computer Assisted Intervention--MICCAI 2019: 22nd International Conference, Shenzhen, China, October...

    2019

  50. [58]

    & Yuille, A

    Xie, L. & Yuille, A. (2017). Genetic cnn. In Proceedings of the IEEE international conference on computer vision (pp.\ 1379--1388)

    2017

  51. [59]

    Xu, Z., Yu, F., Xiong, J., & Chen, X. (2022). Quadralib: A performant quadratic neural network library for architecture optimization and design exploration. Proceedings of Machine Learning and Systems , 4, 503--514

    2022

  52. [60]

    Yan, Z., Yang, X., & Cheng, K.-T. (2018). Joint segment-level and pixel-wise losses for deep learning based retinal vessel segmentation. IEEE Transactions on Biomedical Engineering , 65(9), 1912--1923

    2018

  53. [61]

    Zador, A., Escola, S., Richards, B., \"O lveczky, B., Bengio, Y., Boahen, K., Botvinick, M., Chklovskii, D., Churchland, A., Clopath, C., et al. (2022). Toward next-generation artificial intelligence: Catalyzing the neuroai revolution. arXiv preprint arXiv:2210.08340

    2022 arXiv

  54. [62]

    Zador, A., Escola, S., Richards, B., \"O lveczky, B., Bengio, Y., Boahen, K., Botvinick, M., Chklovskii, D., Churchland, A., Clopath, C., et al. (2023). Catalyzing next-generation artificial intelligence through neuroai. Nature communications , 14(1), 1597

    2023

  55. [63]

    & Klein, J.-C

    Zana, F. & Klein, J.-C. (2001). Segmentation of vessel-like patterns using mathematical morphology and curvature evaluation. IEEE transactions on image processing , 10(7), 1010--1019

    2001

  56. [64]

    Zhang, K., Zhu, D., Min, X., & Zhai, G. (2022). Implicit neural representation learning for hyperspectral image super-resolution. IEEE Transactions on Geoscience and Remote Sensing , 61, 1--12

    2022

  57. [65]

    Zhang, M., Lucas, J., Ba, J., & Hinton, G. E. (2019). Lookahead optimizer: k steps forward, 1 step back. Advances in neural information processing systems , 32

    2019

  58. [66]

    Zhou, Z., Siddiquee, M. M. R., Tajbakhsh, N., & Liang, J. (2019). Unet++: Redesigning skip connections to exploit multiscale features in image segmentation. IEEE transactions on medical imaging , 39(6), 1856--1867

    2019

  59. [67]

    Zoumpourlis, G., Doumanoglou, A., Vretos, N., & Daras, P. (2017). Non-linear convolution filters for cnn-based learning. In Proceedings of the IEEE International Conference on Computer Vision (pp.\ 4761--4769)

    2017

  60. [68]

    , " * write output.state after.block = add.period write newline

    ENTRY address archive author booktitle chapter doi edition editor eid eprint howpublished institution journal key keywords month note number organization pages publisher school series title type url volume year archivePrefix primaryClass adsurl adsnote version label extra.labe...

  61. [69]

    write newline

    " write newline "" before.all 'output.state := FUNCTION add.period duplicate empty 'skip "." * add.blank if FUNCTION if.digit duplicate "0" = swap duplicate "1" = swap duplicate "2" = swap duplicate "3" = swap duplicate "4" = swap duplicate "5" = swap duplicate "6" = swap dupl...

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.