Pith. sign in

REVIEW 4 major objections 6 minor 38 references

UltraDfeGAN: Detail-Enhancing Generative Adversarial Networks for High-Fidelity Functional Ultrasound Synthesis

T0 review · 4 major / 6 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read A GAN with detail feature enhancement and class-conditional batch normalization synthesizes functional ultrasound images that outperform four baselines, and augmenting real training data with them improves downstream classification.

desk verdict Plausible GAN for fUS synthesis, but sloppy task labeling and missing variance make the quantitative claims unverifiable as written. read the letter →

arxiv 2507.03341 v2 pith:WJQHREA7 submitted 2025-07-04 eess.IV cs.CVphysics.med-ph

Zhuo Li , Xuhang Chen , Shuqiang Wang , Bin Yuan , Nou Sotheany , Ngeth Rithea This is my paper · ORCID
classification eess.IVcs.CVphysics.med-ph
keywords functionalultrasoundgenerativeadversarialnetworkimagesynthesisdataaugmentationdetailfeatureenhancementclass-conditionalbatchnormalizationU-Netdiscriminatormedicalimaging
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

Functional ultrasound imaging can watch brain activity in real time, but human brain data are scarce and hard to acquire. This paper argues that a generative adversarial network with two targeted enhancements can synthesize functional ultrasound images that are more faithful than those from diffusion or other leading generative models. The claimed result matters because synthetic images could be used to augment small training sets, and the paper reports that doing so improves a downstream task-state classifier. The central claim is that the proposed architecture produces continuous microvascular structures and sharply defined flow signals that existing baselines miss.

What carries the argument

The core objects are the Detail Feature Enhancement (DFE) module, the class-conditional batch normalization (cBatchNorm) layer, and a U-Net discriminator with two heads. The DFE module processes a feature map through global average pooling and channel-wise max and average pooling branches, fuses these with shared 1x1 convolutions to produce a channel attention map, and applies it by element-wise multiplication followed by a residual connection, which preserves fine vascular detail while suppressing noise. The cBatchNorm layer normalizes features and then modulates them with class-specific scale and shift parameters generated by small MLPs from the class embedding, injecting brain-region or task-state information into the normalization process. The discriminator's global head classifies the whole image through an MLP on the bottleneck features, while its local head produces a per-pixel realness map from the decoder, forcing the generator to optimize both global coherence and local texture realism. These components jointly carry the argument: DFE preserves microvascular texture, cBatchNorm enforces class-conditioned physiological plausibility, and the dual-head discriminator drives pixel-level fidelity.

What would settle it

Re-running the SSIM, MS-SSIM, FID, and downstream classification evaluations with a verified mapping of the dataset's task labels, or checking whether generated vascular networks match expert manual segmentations for continuity, would directly settle whether the reported gains are real.

Watch

Extended reading notes

Core claim

The paper sets out to build a generator that preserves the fine-grained vascular texture and hemodynamic signal that make functional ultrasound images readable. It claims that the proposed UltraDfeGAN—a U-Net based GAN whose generator inserts a Detail Feature Enhancement (DFE) module and whose discriminator uses class-conditional batch normalization plus a dual global/local head—consistently beats DiT, BigGAN, StyleGAN-XL, and VQGAN on reported quantitative metrics. For the piano-playing task it reports SSIM 0.85, MS-SSIM 0.96, and FID 92.95, the last a 51.5% improvement over BigGAN, and it states that on the connect-the-dots task it outperforms all baselines on every reported metric. Qualitatively, it claims the generated images show continuous microvascular structures and sharp flow signals while baselines show vascular discontinuities, noise, or checkerboard artifacts. It further claims that augmenting the real training data with its generated images improves task-state classification accuracy over training on the original data alone.

Load-bearing premise

The reported comparisons rest on the assumption that the public dataset's task labels and train/test partition are applied correctly, and the paper's own switching between 'piano/guitar' and 'connect-the-dots/line-connecting' leaves room for a mislabeling that would undermine the quantitative claims.

Editorial extensions

If this is right

  • Clinical translation of functional ultrasound gains a data-augmentation tool that directly addresses the scarcity of human brain imaging data.
  • Downstream task-state classifiers can improve when trained on real plus synthetic functional ultrasound images, as the paper demonstrates.
  • The reported quality gap over DiT, BigGAN, StyleGAN-XL, and VQGAN suggests that GANs with explicit detail-enhancement modules are a viable alternative to diffusion models for medical image synthesis.
  • The ablation table shows that both the DFE module and cBatchNorm contribute independently to the final accuracy, so future generators can reuse these components in other architectures.

Reading between the lines

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

  • The improvement in downstream classification might stem partly from augmentation-style regularization rather than photorealism, since the reported FID around 93 is far above natural-image levels; readers should not assume the generated images are indistinguishable from real functional ultrasound.
  • The same architecture could transfer to other data-scarce neuroimaging modalities, such as EEG-derived spectrograms or calcium imaging, with the class embedding reinterpreted as the experimental condition.
  • A sharper test of the qualitative claim would compare the microvascular structure of generated images against manually annotated vessel skeletons; the paper's assertions imply such structure should be present.
  • The text's inconsistency between 'piano/guitar' and 'connect-the-dots/line-connecting' suggests the reported metrics should be rechecked on a verified label mapping before the exact numbers are quoted elsewhere.
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

4 major / 6 minor

Summary. The paper proposes UltraDfeGAN, a class-conditional generative adversarial network for synthesizing functional ultrasound (fUS) images from the public Rabut et al. dataset. The architecture combines a U-Net generator and a two-head U-Net discriminator with a Detail Feature Enhancement module and class-conditional batch normalization. The paper reports qualitative and quantitative comparisons against DiT, BigGAN, StyleGAN-XL, and VQGAN, claims consistent superiority across all metrics, and presents a downstream task-state classification experiment where augmentation with generated images is said to improve accuracy. An ablation study is also reported.

Significance. If the reported results are reproducible, the paper would be a useful contribution to fUS data augmentation, which is a real problem given the scarcity of human fUS data. The authors should be credited for using a publicly available dataset, for comparing against several strong modern generative baselines, for a concrete architectural proposal (DFE module, two-head U-Net discriminator, cBatchNorm), and for attempting a downstream utility evaluation. However, the evaluation section as written does not support the strong claims: the task labels and dataset partitions are internally inconsistent, all metrics are reported as single points without variance or significance tests, and the downstream experiment is underspecified. These are load-bearing issues because the central claims are quantitative outperformance and downstream utility.

major comments (4)
  1. [§4.1 vs §4.4, Fig. 3, Table 1] The task labels and partition counts are internally inconsistent. Section 4.1 describes a motor task as "piano playing" with 720 training and 270 test images, and a cognitive task as "Line-connecting" with 1,530 training and 270 test images. Section 4.4 and the caption of Fig. 3 instead evaluate "guitar-playing" and "line-connecting" tasks, and Table 1 reports ablations for "Guitar playing" and "Line-connecting". The text also switches between "piano" and "guitar" and between "connect-the-dots" and "line-connecting" in several places. Because the reported SSIM=0.85, MS-SSIM=0.96, FID=92.95, and all downstream classification numbers cannot be unambiguously assigned to a specific dataset partition, the central quantitative claims are unverifiable. Please correct the labels and state explicitly which classes and which partition counts were actually used.
  2. [§4.4 and Fig. 4] All quantitative results are single-point estimates. No standard deviations, confidence intervals, number of independent runs, or significance tests are reported for SSIM, MS-SSIM, or FID. The referenced Fig. 4, which is supposed to contain the full quantitative comparison, is not reproduced in the text; only selected values are quoted. Consequently, the statement that UltraDfeGAN "consistently outperforms all baselines across all reported metrics" is not supported by the evidence presented. Please provide the full metric table or figure and report means and standard deviations over multiple random seeds, together with an appropriate significance test.
  3. [§4.5] The downstream augmentation experiment is underspecified. It is not stated how many generated images were added, whether the augmented set was class-balanced, how the PCA/Random Forest hyperparameters were selected, or whether the classifier was evaluated on the held-out test partition. The phrase "significant improvement" is used without any statistical test or confidence interval. Since the generator was trained on the same dataset, the paper must also clarify that the downstream test set was not used for training the generator or the classifier. Please specify the full protocol and provide statistical evidence for the claimed improvement.
  4. [§3.4 and §1 Contribution 3] The role of cBatchNorm is ambiguous. Section 3.4 states that cBatchNorm is incorporated within the discriminator, while Section 1 claims that cBatchNorm "inject[s] brain region-specific information directly into the normalization process" of the synthesis. If the conditioning layer is applied only in the discriminator, the generator is only indirectly conditioned through the adversarial loss; this should be clarified explicitly. This matters because the conditioning mechanism is one of the three claimed contributions.
minor comments (6)
  1. [§4.2] Section 4.2 contains the stray text "01 工作总结" and "三、可视化结果", which appear to be leftover template material; this should be removed.
  2. [§3.1, §3.2, §3.3, Fig. 1] The module name is inconsistent: the paper uses "DFE" in most places but "DEF" in Section 3.1, Section 3.3, and the Fig. 1 caption. Please standardize the abbreviation.
  3. [Fig. 1] Fig. 1 contains the typo "DFE Medule"; this should be corrected to "DFE Module".
  4. [§4.1 and Fig. 2] The dimension ordering in Fig. 2 is inconsistent: it lists (C, W, H) in some places and (C, H, W) in others. Please align the notation with the text, which uses (C, H, W).
  5. [§4.3] It is not stated whether the four baseline models were trained with the same learning rate, optimizer settings, and no-augmentation protocol as UltraDfeGAN; please specify the training setup for all models to make the comparison fair and reproducible.
  6. [§5] The conclusion mentions "model fairness" as one of the addressed challenges, but fairness is not defined, studied, or evaluated anywhere in the paper; please either add a relevant discussion or remove the term.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity; claims are benchmarked against external metrics and held-out data, with internal label inconsistencies noted as non-circular verifiability concerns.

full rationale

UltraDfeGAN's derivation chain is self-contained against external benchmarks. The generation quality claims are quantified with SSIM/MS-SSIM/FID computed on a held-out partition of the Rabut et al. dataset and compared with DiT, BigGAN, StyleGAN-XL, and VQGAN; none of these reported values are recoverable by construction from the training objectives in Eqs. (1)-(3). The downstream classification claim is an empirical comparison of a PCA+Random Forest classifier trained on real training images versus training images augmented with generated samples, with accuracy assessed on a real test set; the augmentation improvement is not definitionally equal to any loss or fitted quantity. The paper's self-citations (e.g., Refs. 16-33, including the Wang et al. GAN patent) appear as related-work context and are not used to justify the architecture or to exclude alternatives, so no self-citation chain makes the result forced. The manuscript does contain internal inconsistencies—'piano-playing' vs 'guitar playing'/'Line-connecting' labels (Section 4.1 vs Fig. 3/Table 1), 'DEF' vs 'DFE' module naming, and the stray template text '01 工作总结' in Section 4.2—and the augmentation experiment lacks a duplicate-real-image control; these are correctness/verifiability and experimental-design concerns, not circularity. No equation or fitted parameter is reused as an output prediction, so no circular step is identified.

Assumptions & free parameters 4 free parameters · 6 assumptions · 3 invented entities

The central claims rely on standard domain assumptions about fUS physiology and dataset validity, plus three architectural components that are presented as new but are largely combinations of known techniques. The only true free parameters are hand-chosen hyperparameters and unspecified downstream classifier settings; the model weights themselves are learned and not enumerated here.

free parameters (4)
  • learning_rate = 2e-4
    Hand-chosen Adam learning rate for both generator and discriminator (Section 4.3).
  • beta1, beta2 = 0.5, 0.999
    Hand-chosen Adam momentum parameters (Section 4.3).
  • image_crop_size = 128x128
    All fUS images cropped to 128x128 for efficiency (Section 4.1); affects FID and SSIM comparison.
  • downstream_classifier_hyperparameters = not reported
    The PCA plus Random Forest classifier (Section 4.5) has unspecified number of components and trees, which could affect accuracy numbers.
assumptions (6)
  • domain assumption Neurovascular coupling is a valid basis for inferring neural activity from fUS signals.
    Invoked in the introduction and abstract as the physiological foundation of fUS.
  • domain assumption The Rabut et al. public dataset is a valid and correctly labeled source of fUS images.
    All experiments use this dataset (Section 4.1); the inconsistent task names ('piano' versus 'guitar', 'connect-the-dots' versus 'line-connecting') make this assumption questionable.
  • domain assumption FID and SSIM are appropriate metrics for assessing fUS image fidelity.
    Used as the primary quantitative evaluation in Section 4.4 without justification for their relevance to physiological plausibility.
  • ad hoc to paper The U-Net discriminator design (after Schonfeld et al.) improves adversarial training for this task.
    Adopted as the backbone and credited to reference [38], but the specific benefit for fUS is not independently established.
  • ad hoc to paper The Detail Feature Enhancement module preserves microvascular details via channel attention.
    The module is a variant of SE/CBAM attention; its effect on vascular structure is asserted but not directly measured.
  • ad hoc to paper Class-conditional batch normalization in the discriminator guides the generator to produce class-appropriate images.
    cBatchNorm is placed in the discriminator (Section 3.4), and the paper claims it 'enables the model to synthesize images that accurately reflect the distinct hemodynamic characteristics', but the mechanism is indirect and not demonstrated.
invented entities (3)
  • Detail Feature Enhancement (DFE) module
    purpose: Multi-scale feature fusion and dynamic channel attention to preserve microvascular detail and suppress noise in generated fUS images.
    Introduced as a contribution in Section 3.3, but it is structurally a channel attention module similar to SENet and CBAM, with no external validation beyond the paper's own metrics.
  • Class-Conditional Batch Normalization (cBatchNorm) layer
    purpose: Inject class-specific information into the normalization process in the discriminator to guide the generator toward class-appropriate fUS images.
    Presented as a contribution in Section 3.4, but it is exactly conditional batch normalization from the cBN literature, and its placement in the discriminator makes the claimed synthesis guidance indirect.
  • Two-head U-Net discriminator
    purpose: Perform global and local discrimination simultaneously to encourage globally coherent and locally detailed generated images.
    Combines the U-Net discriminator of Schonfeld et al. with a patch-based local head; a new combination but not a new class of entity.

how reviews work

0 comments
Cite this review

Pith. "Pith review of UltraDfeGAN: Detail-Enhancing Generative Adversarial Networks for High-Fidelity Functional Ultrasound Synthesis." pith.science (2026). https://pith.science/paper/WJQHREA7

@misc{pith2026250703341,
  author       = {Pith},
  title        = {Pith review of: UltraDfeGAN: Detail-Enhancing Generative Adversarial Networks for High-Fidelity Functional Ultrasound Synthesis},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/WJQHREA7}},
  note         = {Machine review of arXiv:2507.03341}
}
read the original abstract

Functional ultrasound (fUS) is a neuroimaging technique known for its high spatiotemporal resolution, enabling non-invasive observation of brain activity through neurovascular coupling. Despite its potential in clinical applications such as neonatal monitoring and intraoperative guidance, the development of fUS faces challenges related to data scarcity and limitations in generating realistic fUS images. This paper explores the use of a generative adversarial network (GAN) framework tailored for fUS image synthesis. The proposed method incorporates architectural enhancements, including feature enhancement modules and normalization techniques, aiming to improve the fidelity and physiological plausibility of generated images. The study evaluates the performance of the framework against existing generative models, demonstrating its capability to produce high-quality fUS images under various experimental conditions. Additionally, the synthesized images are assessed for their utility in downstream tasks, showing improvements in classification accuracy when used for data augmentation. Experimental results are based on publicly available fUS datasets, highlighting the framework's effectiveness in addressing data limitations.

Figures

Figures reproduced from arXiv: 2507.03341 by the authors.

Figure 1
Figure 1. The UltraDfeGAN framework employs a two-UNet architecture as its encoder￾decoder backbone, enhanced through the incorporation of DEF modules and cBatch￾Norm modules. 3.2 UltraDfeGAN Architecture The core of UltraDfeGAN is a generative adversarial network built upon a mod￾ified U-Net architecture for both the generator and the discriminator. Generator. The generator network is designed to transform a random noise vec… view at source ↗
Figure 2
Figure 2. The DEF module architecture diagram presents the integration of multi-scale feature fusion and dynamic attention mechanisms, designed to enhance structural detail preservation and physiological coherence in functional ultrasound imaging. Within each cBatchNorm layer, the feature map x is first normalized using its batch-derived mean µ and standard deviation σ. The normalized activation is then modulated by a class-s… view at source ↗
Figure 3
Figure 3. This figure presents a comparative analysis of generative performance across dif￾ferent models under guitar-playing and Line-connecting tasks in both task and resting￾state scenarios. The first column displays real images, followed by generated results from UltraDfeGAN, DiT, BigGAN, StyleGAN-XL, and VQGAN. 4.3 Experimental Setup We benchmark our proposed model, UltraDfeGAN, against several prominent generative model… view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: A quantitative comparison of image quality metrics across generative models further demonstrates that UltraDfeGAN consistently outperforms all baseline models in terms of quantitative evaluation metrics. The quantitative results, presented in [PITH_FULL_IMAGE:figures/…
Figure 5
Figure 5. Figure 5: Quantitative Evaluation of Downstream Classification Task Performance. 4.6 Ablation Studies To validate the contributions of the core components of UltraDfeGAN, we con￾ducted a series of ablation studies. In these experiments, we systematically re￾verted our proposed a…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

38 extracted references · 34 canonical work pages

  1. [1]

    Nature communications 10(1), 1400 (2019)

    Dizeux, A., Gesnik, M., Ahnine, H., Blaize, K., Arcizet, F., Picaud, S., Sahel, J.A., Deffieux, T., Pouget, P., Tanter, M.: Functional ultrasound imaging of the brain reveals propagation of task-related brain activity in behaving primates. Nature communications 10(1), 1400 (2019)

    work page 2019

  2. [2]

    Nature methods19(8), 1004–1012 (2022)

    Renaudin, N., Demené, C., Dizeux, A., Ialy-Radio, N., Pezet, S., Tanter, M.: Func- tional ultrasound localization microscopy reveals brain-wide neurovascular activity on a microscopic scale. Nature methods19(8), 1004–1012 (2022)

    work page 2022

  3. [3]

    Proceedings of the National Academy of Sciences 117(25), 14453–14463 (2020) Detail-Enhancing Generative Adversarial Networks 11

    Blaize, K., Arcizet, F., Gesnik, M., Ahnine, H., Ferrari, U., Deffieux, T., Pouget, P., Chavane, F., Fink, M., Sahel, J.A., et al.: Functional ultrasound imaging of deep visual cortex in awake nonhuman primates. Proceedings of the National Academy of Sciences 117(25), 14453–14463 (2020) Detail-Enhancing Generative Adversarial Networks 11

    work page 2020

  4. [4]

    Nature Protocols 16(7), 3547–3571 (2021)

    Brunner, C., Grillet, M., Urban, A., Roska, B., Montaldo, G., Macé, E.: Whole- brain functional ultrasound imaging in awake head-fixed mice. Nature Protocols 16(7), 3547–3571 (2021)

    work page 2021

  5. [5]

    Nature communications 12(1), 1080 (2021)

    Baranger, J., Demene, C., Frerot, A., Faure, F., Delanoë, C., Serroune, H., Houdouin, A., Mairesse, J., Biran, V., Baud, O., et al.: Bedside functional monitor- ing of the dynamic brain connectivity in human neonates. Nature communications 12(1), 1080 (2021)

    work page 2021

  6. [6]

    Science translational medicine9(411), eaah6756 (2017)

    Demene,C.,Baranger,J.,Bernal,M.,Delanoe,C.,Auvin,S.,Biran,V.,Alison,M., Mairesse, J., Harribaud, E., Pernot, M., et al.: Functional ultrasound imaging of brain activity in human newborns. Science translational medicine9(411), eaah6756 (2017)

    work page 2017

  7. [7]

    In: 2016 IEEE International Ultrasonics Symposium (IUS)

    Demene, C., Bernal, M., Delanoe, C., Auvin, S., Biran, V., Alison, M., Mairesse, J., Harribaud, E., Pernot, M., Tanter, M., et al.: Functional ultrasound imaging of the brain activity in human neonates. In: 2016 IEEE International Ultrasonics Symposium (IUS). pp. 1–3. IEEE (2016)

    work page 2016

  8. [8]

    Imbault, M., Chauvet, D., Gennisson, J.L., Capelle, L., Tanter, M.: Intraoperative functionalultrasoundimagingofhumanbrainactivity.Scientificreports 7(1), 7304 (2017)

    work page 2017

Show all 38 references
  1. [9]

    Frontiers in neuroscience13, 1384 (2020)

    Soloukey, S., Vincent, A.J., Satoer, D.D., Mastik, F., Smits, M., Dirven, C.M., Strydis, C., Bosch, J.G., van der Steen, A.F., De Zeeuw, C.I., et al.: Functional ul- trasound (fus) during awake brain surgery: the clinical potential of intra-operative functional and vascular br...

    2020

  2. [10]

    Soloukey, S., Verhoef, L., Mastik, F., Brown, M., Generowicz, B., Satoer, D., Dirven, C., Smits, M., Koekkoek, S., Vincent, A., et al.: P09. 02. aa window to the brain: Monitoring the human brain using functional ultrasound (fus)-imaging in freely moving subjects with an acous...

    2024

  3. [11]

    Nature Neuroscience27(1), 196–207 (2024)

    Griggs, W.S., Norman, S.L., Deffieux, T., Segura, F., Osmanski, B.F., Chau, G., Christopoulos, V., Liu, C., Tanter, M., Shapiro, M.G., et al.: Decoding motor plans using a closed-loop ultrasonic brain–machine interface. Nature Neuroscience27(1), 196–207 (2024)

    2024

  4. [12]

    Re- search 6, 0200 (2023)

    Zheng, H., Niu, L., Qiu, W., Liang, D., Long, X., Li, G., Liu, Z., Meng, L.: The emergence of functional ultrasound for noninvasive brain–computer interface. Re- search 6, 0200 (2023)

    2023

  5. [13]

    Neuron112(10), 1710–1722 (2024)

    Agyeman, K.A., Lee, D.J., Russin, J., Kreydin, E.I., Choi, W., Abedi, A., Lo, Y.T., Cavaleri, J., Wu, K., Edgerton, V.R., et al.: Functional ultrasound imaging of the human spinal cord. Neuron112(10), 1710–1722 (2024)

    2024

  6. [14]

    Advances in neural in- formation processing systems27 (2014)

    Goodfellow, I.J., Pouget-Abadie, J., Mirza, M., Xu, B., Warde-Farley, D., Ozair, S., Courville, A., Bengio, Y.: Generative adversarial nets. Advances in neural in- formation processing systems27 (2014)

    2014

  7. [15]

    Science Trans- lational Medicine 16(749), eadj3143 (2024)

    Rabut, C., Norman, S.L., Griggs, W.S., Russin, J.J., Jann, K., Christopoulos, V., Liu, C., Andersen, R.A., Shapiro, M.G.: Functional ultrasound imaging of human brain activity through an acoustically transparent cranial window. Science Trans- lational Medicine 16(749), eadj3143 (2024)

    2024

  8. [16]

    Brain imaging and behavior15, 276–287 (2021)

    Lei, B., Yu, S., Zhao, X., Frangi, A.F., Tan, E.L., Elazab, A., Wang, T., Wang, S.: Diagnosis of early alzheimer’s disease based on dynamic high order networks. Brain imaging and behavior15, 276–287 (2021)

    2021

  9. [17]

    Wang, S., Yu, W., Chen, Z., et al.: Smart diagnosis assistance method to solve results of inaccurate classification of image, and terminal based on medical images (Mar 18 2025), uS Patent 12,254,684 12 Zhuo Li et al

    2025

  10. [18]

    In: 2018 International conference on artificial intelligence and big data (ICAIBD)

    Wang, S., Shen, Y., Zeng, D., Hu, Y.: Bone age assessment using convolutional neural networks. In: 2018 International conference on artificial intelligence and big data (ICAIBD). pp. 175–178. IEEE (2018)

    2018

  11. [19]

    Scientific Reports13(1), 12098 (2023)

    Müller-Franzes, G., Niehues, J.M., Khader, F., Arasteh, S.T., Haarburger, C., Kuhl, C., Wang, T., Han, T., Nolte, T., Nebelung, S., et al.: A multimodal compar- ison of latent denoising diffusion probabilistic models and generative adversarial networks for medical image synthe...

    2023

  12. [20]

    Procedia computer science111, 17–23 (2017)

    Zeng, D., Wang, S., Shen, Y., Shi, C.: A ga-based feature selection and parameter optimization for support tucker machine. Procedia computer science111, 17–23 (2017)

    2017

  13. [21]

    Mo, L.F., Wang, S.Q.: A variational approach to nonlinear two-point boundary valueproblems.NonlinearAnalysis:Theory,Methods&Applications 71(12),e834– e838 (2009)

    2009

  14. [22]

    Wang, S., Yanyan, S., Zhang, W.: Enhanced generative adversarial network and target sample recognition method (Nov 26 2024), uS Patent 12,154,036

    2024

  15. [23]

    Deep learning applications pp

    Wang, S., Wang, H., Cheung, A.C., Shen, Y., Gan, M.: Ensemble of 3d densely connected convolutional network for diagnosis of mild cognitive impairment and alzheimer’s disease. Deep learning applications pp. 53–73 (2020)

    2020

  16. [24]

    IEEE Transactions on Circuits and Systems for Video Technology (2025)

    Guo, X., Chen, X., Wang, S., Pun, C.M.: Underwater image restoration through a prior guided hybrid sense approach and extensive benchmark analysis. IEEE Transactions on Circuits and Systems for Video Technology (2025)

    2025

  17. [25]

    IEEE Transactions on Neural Systems and Rehabilitation Engineering31, 4601–4612 (2023)

    Zuo, Q., Shen, Y., Zhong, N., Chen, C.P., Lei, B., Wang, S.: Alzheimer’s disease prediction via brain structural-functional deep fusing network. IEEE Transactions on Neural Systems and Rehabilitation Engineering31, 4601–4612 (2023)

    2023

  18. [26]

    IEEE Transactions on Cybernetics54(6), 3652–3665 (2024)

    Zuo, Q., Wu, H., Chen, C.P., Lei, B., Wang, S.: Prior-guided adversarial learning with hypergraph for predicting abnormal connections in alzheimer’s disease. IEEE Transactions on Cybernetics54(6), 3652–3665 (2024)

    2024

  19. [27]

    In: Pattern Recognition and Computer Vision: 4th Chinese Conference, PRCV 2021, Beijing, China, October 29–November 1, 2021, Proceedings, Part III 4

    Pan, J., Lei, B., Shen, Y., Liu, Y., Feng, Z., Wang, S.: Characterization multimodal connectivity of brain network by hypergraph gan for alzheimer’s disease analysis. In: Pattern Recognition and Computer Vision: 4th Chinese Conference, PRCV 2021, Beijing, China, October 29–Nov...

    2021

  20. [28]

    In: Medical imaging with deep learning (2022)

    Dorjsembe, Z., Odonchimed, S., Xiao, F.: Three-dimensional medical image syn- thesis with denoising diffusion probabilistic models. In: Medical imaging with deep learning (2022)

    2022

  21. [29]

    In: Pattern Recognition and Computer Vision: 4th Chinese Conference, PRCV 2021, Beijing, China, October 29–November 1, 2021, Proceedings, Part III 4

    Zuo, Q., Lei, B., Shen, Y., Liu, Y., Feng, Z., Wang, S.: Multimodal representations learning and adversarial hypergraph fusion for early alzheimer’s disease prediction. In: Pattern Recognition and Computer Vision: 4th Chinese Conference, PRCV 2021, Beijing, China, October 29–N...

    2021

  22. [30]

    IEEE Transactions on Pattern Analysis and Machine Intelligence (2024)

    Zong, Y., Zuo, Q., Ng, M.K.P., Lei, B., Wang, S.: A new brain network construction paradigm for brain disorder via diffusion-based graph contrastive learning. IEEE Transactions on Pattern Analysis and Machine Intelligence (2024)

    2024

  23. [31]

    Li, Y., Liao, I.Y., Zhong, N., Toshihiro, F., Wang, Y., Wang, S.: Generative ai en- ablesthedetectionofautismusingeegsignals.In:ChineseConferenceonBiometric Recognition. pp. 375–384. Springer Nature Singapore Singapore (2023)

    2023

  24. [32]

    IEEE Transactions on Cybernetics54(9), 5026–5039 (2024) Detail-Enhancing Generative Adversarial Networks 13

    Jing, C., Shen, Y., Zhao, S., Pan, Y., Chen, C.P., Lei, B., Wang, S.: Estimating addiction-related brain connectivity by prior-embedding graph generative adver- sarial networks. IEEE Transactions on Cybernetics54(9), 5026–5039 (2024) Detail-Enhancing Generative Adversarial Networks 13

    2024

  25. [33]

    IEEE Transactions on Neural Systems and Rehabilitation Engineering31, 4017– 4028 (2023)

    Zuo, Q., Zhong, N., Pan, Y., Wu, H., Lei, B., Wang, S.: Brain structure-function fusing representation learning using adversarial decomposed-vae for analyzing mci. IEEE Transactions on Neural Systems and Rehabilitation Engineering31, 4017– 4028 (2023)

    2023

  26. [34]

    Peebles,W.,Xie,S.:Scalablediffusionmodelswithtransformers.In:Proceedingsof the IEEE/CVF international conference on computer vision. pp. 4195–4205 (2023)

    2023

  27. [35]

    arXiv preprint arXiv:1809.11096 (2018)

    Brock, A., Donahue, J., Simonyan, K.: Large scale gan training for high fidelity natural image synthesis. arXiv preprint arXiv:1809.11096 (2018)

    2018 arXiv

  28. [36]

    In: ACM SIGGRAPH 2022 conference proceedings

    Sauer, A., Schwarz, K., Geiger, A.: Stylegan-xl: Scaling stylegan to large diverse datasets. In: ACM SIGGRAPH 2022 conference proceedings. pp. 1–10 (2022)

    2022

  29. [37]

    arXiv preprint arXiv:2110.04627 (2021)

    Yu, J., Li, X., Koh, J.Y., Zhang, H., Pang, R., Qin, J., Ku, A., Xu, Y., Baldridge, J., Wu, Y.: Vector-quantized image modeling with improved vqgan. arXiv preprint arXiv:2110.04627 (2021)

    2021 arXiv

  30. [38]

    In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition

    Schonfeld, E., Schiele, B., Khoreva, A.: A u-net based discriminator for generative adversarial networks. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. pp. 8207–8216 (2020)

    2020

Pith tools

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