Pith. sign in

REVIEW 4 major objections 4 minor 29 references

Cross-channel Perception Learning for H&E-to-IHC Virtual Staining

T0 review · 4 major / 4 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read Modeling the nucleus-membrane correlation during H&E-to-IHC translation yields virtual HER2 images that preserve pathologist-scored staining features better than earlier methods.

desk verdict Plausible loss combination and honest ablations, but the paper's own Table I contradicts the SOTA claim—must be corrected before the results can be trusted. read the letter →

arxiv 2506.07559 v1 pith:CPCO7VJN submitted 2025-06-09 cs.CV

Hao Yang , JianYu Wu , Run Fang , Xuelian Zhao , Yuan Ji , Zhiyu Chen , Guibin He , Junceng Guo
show 2 more authors
Yang Liu Xinhua Zeng
This is my paper · ORCID
classification cs.CV
keywords virtualstainingH&E-to-IHCHER2cross-channelcorrelationfeaturedistillationpathologyfoundationmodelfocalopticaldensitydigital
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 seeks to establish that the limited fidelity of H&E-to-IHC virtual staining comes in part from ignoring the relationship between cell nuclei and cell membranes, and that a loss function explicitly encoding that relationship fixes it. CCPL decomposes real HER2 images into hematoxylin (nucleus) and DAB (membrane) channels, uses a pathology foundation model's tile encoder to compare generated and real channels, and adds three losses: dual-channel perception on focal optical density statistics, feature distillation, and nucleus-membrane cross-channel correlation. On the BCI and MIST-HER2 benchmarks, CCPL reports higher pixel-level metrics and higher pathologist ratings for positive-cell proportion and staining-intensity accuracy than the compared methods. If the claim is right, virtual HER2 staining can reproduce the features pathologists actually score, which matters because chemical IHC staining is slow and costly and digital pathology would benefit from reliable surrogate stains.

What carries the argument

The load-bearing object is the Nucleus-Membrane Cross-Channel Correlation (NMCC) loss. For a generated or real HER2 tile, CCPL first uses color deconvolution to isolate the hematoxylin and DAB channels and computes their separate Gigapath features; the correlation between the two channels is then summarized by a single scalar per image, $R = \gamma \|f_H - f_D\|_2^2 + (1-\gamma)(1-\cos(f_H, f_D))$, combining an L2 distance and a cosine term. The NMCC loss compares the generated image's $R_g$ with the real image's $R_r$ in the same mixed L2-plus-cosine form. This scalar is what forces the generator to reproduce the spatial and intensity relationship between nuclei and membranes rather than matching each channel independently. The same Gigapath features also feed the feature-distillation loss, so one frozen encoder carries both the semantic-alignment and the cross-channel targets.

What would settle it

A concrete check: take a set of real HER2 images with pathologist-assigned membrane-intensity and positive-cell-proportion scores, compute $R$ for each, and see whether $R$ orders the images by score. If $R$ is uncorrelated with the scores, the NMCC target is not biologically meaningful. A complementary ablation would remove only the NMCC loss and have pathologists blind-compare the outputs; if the scores do not move, the cross-channel term is not the source of the reported fidelity gain.

Watch

Extended reading notes

Core claim

The central claim, stated on the paper's own terms, is that cross-channel semantic consistency between the nucleus and the membrane is a learnable target for H&E-to-HER2 translation. CCPL obtains single-channel images by color deconvolution and focal optical density conversion, extracts features $f_{H,g}, f_{D,g}$ from the generated image and $f_{H,r}, f_{D,r}$ from the real image using Gigapath's tile encoder, and defines the cross-channel correlation $R = \gamma \|f_H - f_D\|_2^2 + (1-\gamma)(1 - \cos(f_H, f_D))$ for each image. The nucleus-membrane cross-channel loss then enforces $R_g \approx R_r$, alongside a feature-distillation loss that aligns generated and real representations in the same encoder space and a dual-channel perception loss that matches focal optical density statistics. The paper argues that this combination preserves the proportion of HER2-positive cells and the membrane staining intensity that HER2 scoring depends on, and reports quantitative and pathologist-evaluation evidence that it does so better than the compared prior methods.

Load-bearing premise

The argument has a single load-bearing premise: the feature-space difference between the hematoxylin and DAB channels, condensed into $R$, corresponds to the spatial and intensity relationship between nuclei and membranes that a pathologist reads; if that correspondence fails, the NMCC loss is aligning the generated and real images on a quantity unrelated to diagnosis.

Editorial extensions

If this is right

  • If CCPL is correct, virtual HER2 images will preserve the two features pathologists score\u2014membrane intensity and positive-cell proportion\u2014more faithfully than earlier translation methods.
  • Because the extra losses use a frozen foundation-model encoder at training time only, the deployed generator remains lightweight and fast.
  • Each component has a distinct reported role: feature distillation improves distribution-level realism, NMCC improves structural correlation, and dual-channel perception improves pixel-level detail.
  • The method is demonstrated on two breast-cancer H&E-to-IHC benchmarks, so the gains are not tied to a single dataset.

Reading between the lines

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

  • Editorial inference: the same NMCC loss could be carried over to other membrane-localized IHC markers, such as EGFR or CD20, where positive membranes wrap around nuclei, because the loss itself does not depend on HER2-specific biology.
  • Editorial inference: the reported FID drop when NMCC is added suggests a metric-level ambiguity; a pathology-specific distribution metric would tell whether NMCC genuinely hurts distribution-level realism or only Inception-based FID.
  • Editorial inference: the color deconvolution thresholds and exponents ($T_H, T_D, \alpha_H, \alpha_D$) are fixed hyperparameters; testing CCPL across scanners and antibody kits would show whether the decomposition, and therefore the correlation target, stays stable.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 4 minor

Summary. The paper proposes Cross-Channel Perception Learning (CCPL) for H&E-to-IHC virtual staining. The method decomposes HER2 immunohistochemistry into hematoxylin (nucleus) and DAB (membrane) channels, and introduces three auxiliary losses: a Dual-Channel Perception (DCP) loss based on Focal Optical Density statistics, a Feature Distillation (FD) loss that aligns generated and real image features extracted by Gigapath's Tile Encoder, and a Nucleus-Membrane Cross-Channel Correlation (NMCC) loss. Experiments are reported on the BCI and MIST-HER2 datasets using PSNR, SSIM, PCC, FID, and pathologist ratings, with the paper claiming state-of-the-art performance.

Significance. If the claims were fully supported, the idea of exploiting cross-channel nucleus-membrane correlations with a pathology foundation model would be a useful contribution to virtual staining, and the lightweight distillation approach is practically attractive. The paper is commendable for giving explicit loss definitions, using ground-truth-supervised objectives, and including an ablation study with human evaluation. However, the central quantitative claim is contradicted by the paper's own Table I, and the NMCC loss as written is technically questionable, so the significance as currently stated cannot be accepted without revision.

major comments (4)
  1. [Section IV-C, Table I] The text states that "CCPL achieves the lowermost FID scores on the BCI dataset," but Table I reports FID 57.81 for Ours versus 48.42 for PSPStain and 54.72 for ASP; since lower FID is better, this statement is false. The abstract and contribution list also claim state-of-the-art PCC, but Table I shows ASP has PCC 0.2609 versus 0.2441 for Ours on BCI. The central claim that CCPL achieves state-of-the-art quantitative performance is therefore not supported by the paper's own results, and the affected statements must be corrected and the claims appropriately qualified.
  2. [Section III-C, Eqs. (7)-(9)] R_g and R_r are defined as scalars in Eqs. (7) and (8): each is a weighted sum of an L2 norm and a cosine distance, both of which are scalar quantities. Eq. (9) then computes an L2 norm and a cosine similarity between R_g and R_r. A cosine similarity between scalars is degenerate (it can only be +1 or -1 depending on sign), so the second term of L_cross is either zero or a constant, and the loss reduces to a scalar L2 distance. This makes the proposed cross-channel consistency loss poorly motivated and its gradient behavior unclear. The authors should either define R as a vector-valued correlation representation or justify why a scalar distance is sufficient, and provide an analysis of the loss landscape.
  3. [Section IV-A, Human Evaluation] The human evaluation protocol is not reproducible as described. The text states that "images with significant scoring discrepancies were excluded" but does not specify the discrepancy threshold, the number of excluded images, the number of images scored, or how the three pathologists' scores were aggregated. In addition, the human evaluation is cited to reference [17], which is an arXiv paper on social conversational dialogues and appears unrelated to pathology or image evaluation. Since human evaluation is part of the paper's central claim, full details and an appropriate citation are required.
  4. [Section III-C] The NMCC premise that the feature-space distance and cosine similarity between Gigapath features of the hematoxylin and DAB channels capture the biologically meaningful nucleus-membrane relationship is not validated. The paper does not show that minimizing the difference between R_generated and R_real actually enforces spatial or intensity correlations that are pathologically meaningful; it could simply be aligning arbitrary feature statistics. A concrete validation, such as correlation with HER2 scoring annotations or a comparison against a simpler cross-channel pixel-space alignment, is needed to support the central contribution.
minor comments (4)
  1. [Eq. (1)] The notation in Eq. (1) is confusing: "F ODC = ((ODC)αC, if F O D C > T_C, 0 otherwise" appears to use F ODC both as the output and in the condition; it should presumably be ODC > T_C. Please clarify.
  2. [Throughout] There are several typographical errors, including "demote" for "denote" (Eq. (2)), "groud truth" (Fig. 2 caption), "evlautions" and "bechmark" in the abstract/conclusion areas, and "Mode 3 vs. Model 4" in Section IV-D that should be "Model 3 vs. Model 4".
  3. [Section IV-D, Table III] The ablation discussion attributes the FID decline for NMCC to the Inception feature extractor, but the full model still has a much higher FID (57.81) than the FD-only model (40.05), even though FD-only has the best FID overall. The text should more clearly acknowledge that the final model is not FID-optimal and explain this trade-off with respect to the SOTA FID claim in Section IV-C.
  4. [Section IV-A] The definitions of PCPA and SIA should state exactly which HER2 scoring guideline is used and how the "proportion of positive cell nuclei" is determined, since HER2 positivity is based on membrane staining rather than nuclear positivity in the standard clinical scoring.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity; losses are direct supervision against real IHC images and no fitted parameter is renamed as a prediction.

full rationale

I found no circular step. CCPL's losses (DCP Eq. (2), FD Eqs. (4)-(6), NMCC Eqs. (7)-(9)) all compare quantities computed from the generated image against the same quantities computed from the paired real IHC image; they are direct supervision terms, not derived predictions. The FOD thresholds T_H/T_D and the alpha/beta/gamma weights are hand-set hyperparameters (Section IV-B) and are not fitted to the test set and then reported as predictions. Gigapath's Tile Encoder is a frozen external foundation model, so the feature distillation is not self-referential. The ablation study (Table III) tests each module against the full model and does not redefine metrics. The text's claim that CCPL achieves 'lowermost FID' on BCI conflicts with Table I (Ours 57.81 vs ASP 54.72, PSPStain 48.42), and the PCC claim conflicts with Table I (ASP 0.2609 vs Ours 0.2441 on BCI); this is an internal consistency or correctness problem, not circularity. The citation [17] to the authors' prior Stephaine paper is unrelated to the human evaluation and is an inappropriate or incorrect citation, but it is not load-bearing: the actual human evaluation is conducted and reported in this paper, so no derivation reduces to a self-citation.

Assumptions & free parameters 6 free parameters · 4 assumptions · 0 invented entities

The method introduces no new physical entities. Its load-bearing assumptions are domain assumptions about stain-channel semantics and the meaningfulness of Gigapath feature distances for cross-channel correlation. The free parameters are hand-set hyperparameters, and several are not reported, which hurts reproducibility but does not introduce circular fitting.

free parameters (6)
  • FOD thresholds T_H, T_D = 0.15
    Global thresholds used to remove background noise in Eq. 1; hand-chosen in Section IV-B.
  • FOD alpha coefficients alpha_H, alpha_D = 1.8
    Control the degree of focusing on optical density values in Eq. 1; hand-chosen in Section IV-B.
  • Dual-channel balance weight alpha = 0.1
    Balances hematoxylin and DAB losses in Eq. 3; hand-chosen in Section IV-B.
  • Feature distillation weight beta = not reported
    Balances cosine and L2 distillation losses in Eq. 6; no value is specified.
  • Cross-channel balance weight gamma = not reported
    Balances L2 and cosine terms in cross-channel correlation R in Eqs. 7-8; no value is specified.
  • Cross-channel consistency weight theta = not reported
    Balances L2 and cosine terms in cross-channel consistency loss in Eq. 9; no value is specified.
assumptions (4)
  • domain assumption Color deconvolution with the HED matrix separates hematoxylin and DAB channels reliably.
    Used in Section III-A to split the HER2 image into nucleus and membrane channels; relies on standard stain separation assumptions.
  • domain assumption The DAB channel corresponds to cell membranes and hematoxylin to cell nuclei in HER2 staining.
    Basis for dual-channel perception; if incorrect, the perceptual losses align the wrong structures.
  • domain assumption Gigapath's tile encoder produces features that capture pathology-relevant semantics for the correlation measurement.
    Used in Section III-B and III-C to compute distillation and cross-channel losses; the pretrained encoder is treated as a reliable semantic extractor.
  • domain assumption Focal Optical Density (FOD) with the given thresholds emphasizes tumor-relevant regions.
    Adopted from [14] and applied to both channels; the alpha and thresholds are hand-set and not independently validated.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Cross-channel Perception Learning for H&E-to-IHC Virtual Staining." pith.science (2026). https://pith.science/paper/CPCO7VJN

@misc{pith2026250607559,
  author       = {Pith},
  title        = {Pith review of: Cross-channel Perception Learning for H&E-to-IHC Virtual Staining},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/CPCO7VJN}},
  note         = {Machine review of arXiv:2506.07559}
}
read the original abstract

With the rapid development of digital pathology, virtual staining has become a key technology in multimedia medical information systems, offering new possibilities for the analysis and diagnosis of pathological images. However, existing H&E-to-IHC studies often overlook the cross-channel correlations between cell nuclei and cell membranes. To address this issue, we propose a novel Cross-Channel Perception Learning (CCPL) strategy. Specifically, CCPL first decomposes HER2 immunohistochemical staining into Hematoxylin and DAB staining channels, corresponding to cell nuclei and cell membranes, respectively. Using the pathology foundation model Gigapath's Tile Encoder, CCPL extracts dual-channel features from both the generated and real images and measures cross-channel correlations between nuclei and membranes. The features of the generated and real stained images, obtained through the Tile Encoder, are also used to calculate feature distillation loss, enhancing the model's feature extraction capabilities without increasing the inference burden. Additionally, CCPL performs statistical analysis on the focal optical density maps of both single channels to ensure consistency in staining distribution and intensity. Experimental results, based on quantitative metrics such as PSNR, SSIM, PCC, and FID, along with professional evaluations from pathologists, demonstrate that CCPL effectively preserves pathological features, generates high-quality virtual stained images, and provides robust support for automated pathological diagnosis using multimedia medical data.

Figures

Figures reproduced from arXiv: 2506.07559 by the authors.

Figure 1
Figure 1. Overview of the Cross-channel Perception. (a) Real and generated images are passed through Gigapath’s tile encoder to extract features, with the feature [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Illustration of virtually stained results of the CCPL, ASP [27], PSPStain [14], Pyramidpix2pix [26] and Pix2Pix [25] on selected samples form the [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

29 extracted references · 28 canonical work pages

  1. [17]

    Stephanie: Step-by-Step Dialogues for Mimicking Human Interactions in Social Conversations

    Hao Yang, Hongyuan Lu, Xinhua Zeng, Yang Liu, Xiang Zhang, Haoran Yang, Yumeng Zhang, Shan Huang, Yiran Wei, and Wai Lam, “Stephanie: Step-by-step dialogues for mimicking human interactions in social conversations,”arXiv preprint arXiv:2407.04093, 2024

    work page Pith review arXiv 2024

  2. [1]

    Interaction-matrix based personalized image aesthetics assessment,

    Jingwen Hou, Weisi Lin, Guanghui Yue, Weide Liu, and Baoquan Zhao, “Interaction-matrix based personalized image aesthetics assessment,” IEEE Transactions on Multimedia, vol. 25, pp. 5263–5278, 2022

    work page 2022

  3. [2]

    Cross- image region mining with region prototypical network for weakly supervised segmentation,

    Weide Liu, Xiangfei Kong, Tzu-Yi Hung, and Guosheng Lin, “Cross- image region mining with region prototypical network for weakly supervised segmentation,”IEEE Transactions on Multimedia, vol. 25, pp. 1148–1160, 2021

    work page 2021

  4. [3]

    Decoding bilingual eeg signals with complex semantics using adaptive graph attention convolutional network,

    Chengfang Li, Yang Liu, Jielin Li, Yuhao Miao, Jing Liu, and Liang Song, “Decoding bilingual eeg signals with complex semantics using adaptive graph attention convolutional network,”IEEE Transactions on Neural Systems and Rehabilitation Engineering, 2024

    work page 2024

  5. [4]

    Few-shot segmentation with optimal transport matching and message flow,

    Weide Liu, Chi Zhang, Henghui Ding, Tzu-Yi Hung, and Guosheng Lin, “Few-shot segmentation with optimal transport matching and message flow,”IEEE Transactions on Multimedia, vol. 25, pp. 5130–5141, 2022

    work page 2022

  6. [5]

    Deep learning- enabled realistic virtual histology with ultraviolet photoacoustic remote sensing microscopy,

    Matthew T Martell, Nathaniel JM Haven, Brendyn D Cikaluk, Brendon S Restall, Ewan A McAlister, Rohan Mittal, Benjamin A Adam, Nadia Giannakopoulos, Lashan Peiris, Sveta Silverman, et al., “Deep learning- enabled realistic virtual histology with ultraviolet photoacoustic remote sensing microscopy,”Nature Communications, vol. 14, no. 1, pp. 5967, 2023

    work page 2023

  7. [6]

    Virtual staining for histology by deep learning,

    Leena Latonen, Sonja Koivukoski, Umair Khan, and Pekka Ruusuvuori, “Virtual staining for histology by deep learning,”Trends in Biotechnol- ogy, 2024

    work page 2024

  8. [7]

    Virtual immunohistochemistry staining for histological images assisted by weakly-supervised learning,

    Jiahan Li, Jiuyang Dong, Shenjin Huang, Xi Li, Junjun Jiang, Xiaopeng Fan, and Yongbing Zhang, “Virtual immunohistochemistry staining for histological images assisted by weakly-supervised learning,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2024, pp. 11259–11268

    work page 2024

Show all 29 references
  1. [8]

    Weakly supervised segmentation with maximum bipartite graph matching,

    Weide Liu, Chi Zhang, Guosheng Lin, Tzu-Yi Hung, and Chunyan Miao, “Weakly supervised segmentation with maximum bipartite graph matching,” inProceedings of the 28th ACM International Conference on Multimedia, 2020, pp. 2085–2094

    2020

  2. [9]

    Learning causality-inspired representation consistency for video anomaly detec- tion,

    Yang Liu, Zhaoyang Xia, Mengyang Zhao, Donglai Wei, Yuzheng Wang, Liu Siao, Bobo Ju, Gaoyun Fang, Jing Liu, and Liang Song, “Learning causality-inspired representation consistency for video anomaly detec- tion,” inProceedings of the 31st ACM International Conference on Multime...

    2023

  3. [10]

    A novel cell contour-based instance segmentation model and its ap- plications in her2 breast cancer discrimination,

    Bobo Ju, Yang Liu, Xing Hu, Dongfang Zhao, and Linhua Jiang, “A novel cell contour-based instance segmentation model and its ap- plications in her2 breast cancer discrimination,”Biomedical Signal Processing and Control, vol. 85, pp. 104941, 2023

    2023

  4. [11]

    Deep learning-enabled virtual histological staining of biological samples,

    Bijie Bai, Xilin Yang, Yuzhu Li, Yijie Zhang, Nir Pillar, and Aydogan Ozcan, “Deep learning-enabled virtual histological staining of biological samples,”Light: Science & Applications, vol. 12, no. 1, pp. 57, 2023

    2023

  5. [12]

    Her2 status in breast cancer: changes in guidelines and complicating factors for interpretation,

    Soomin Ahn, Ji Won Woo, Kyoungyul Lee, and So Yeon Park, “Her2 status in breast cancer: changes in guidelines and complicating factors for interpretation,”Journal of pathology and translational medicine, vol. 54, no. 1, pp. 34–44, 2020

    2020

  6. [13]

    Ex- ploiting supervision information in weakly paired images for ihc virtual staining,

    Yueheng Li, Xianchao Guan, Yifeng Wang, and Yongbing Zhang, “Ex- ploiting supervision information in weakly paired images for ihc virtual staining,” inInternational Conference on Medical Image Computing and Computer-Assisted Intervention. Springer, 2024, pp. 113–122

    2024

  7. [14]

    Pathological semantics-preserving learning for h&e-to- ihc virtual staining,

    Fuqiang Chen, Ranran Zhang, Boyun Zheng, Yiwen Sun, Jiahui He, and Wenjian Qin, “Pathological semantics-preserving learning for h&e-to- ihc virtual staining,” inInternational Conference on Medical Image Computing and Computer-Assisted Intervention. Springer, 2024, pp. 384–394

    2024

  8. [15]

    Efficient supervised pretraining of swin- transformer for virtual staining of microscopy images,

    Jiabo Ma and Hao Chen, “Efficient supervised pretraining of swin- transformer for virtual staining of microscopy images,”IEEE Transac- tions on Medical Imaging, 2023

    2023

  9. [16]

    A whole-slide foundation model for digital pathology from real-world data,

    Hanwen Xu, Naoto Usuyama, Jaspreet Bagga, Sheng Zhang, Rajesh Rao, Tristan Naumann, Cliff Wong, Zelalem Gero, Javier Gonz ´alez, Yu Gu, et al., “A whole-slide foundation model for digital pathology from real-world data,”Nature, pp. 1–8, 2024

    2024

  10. [18]

    Generalized video anomaly event detection: Systematic taxonomy and comparison of deep models,

    Yang Liu, Dingkang Yang, Yan Wang, Jing Liu, Jun Liu, Azzedine Boukerche, Peng Sun, and Liang Song, “Generalized video anomaly event detection: Systematic taxonomy and comparison of deep models,” ACM Computing Surveys, vol. 56, no. 7, pp. 1–38, 2024

    2024

  11. [19]

    Deepdof-se: affordable deep-learning microscopy platform for slide-free histology,

    Lingbo Jin, Yubo Tang, Jackson B Coole, Melody T Tan, Xuan Zhao, Hawraa Badaoui, Jacob T Robinson, Michelle D Williams, Nadarajah Vigneswaran, Ann M Gillenwater, et al., “Deepdof-se: affordable deep-learning microscopy platform for slide-free histology,”Nature Communications, ...

    2024

  12. [20]

    Instant diagnosis of gastroscopic biopsy via deep-learned single-shot femtosecond stimulated raman histology,

    Zhijie Liu, Wei Su, Jianpeng Ao, Min Wang, Qiuli Jiang, Jie He, Hua Gao, Shu Lei, Jinshan Nie, Xuefeng Yan, et al., “Instant diagnosis of gastroscopic biopsy via deep-learned single-shot femtosecond stimulated raman histology,”Nature communications, vol. 13, no. 1, pp. 4050, 2022

    2022

  13. [21]

    Generative adversarial networks in digital histopathology: current applications, limitations, ethical considerations, and future directions,

    Shahd A Alajaji, Zaid H Khoury, Mohamed Elgharib, Mamoon Saeed, Ahmed RH Ahmed, Mohammad B Khan, Tiffany Tavares, Maryam Jessri, Adam C Puche, Hamid Hoorfar, et al., “Generative adversarial networks in digital histopathology: current applications, limitations, ethical consider...

    2024

  14. [22]

    Amp-net: Appearance- motion prototype network assisted automatic video anomaly detection system,

    Yang Liu, Jing Liu, Kun Yang, Bobo Ju, Siao Liu, Yuzheng Wang, Dingkang Yang, Peng Sun, and Liang Song, “Amp-net: Appearance- motion prototype network assisted automatic video anomaly detection system,”IEEE Transactions on Industrial Informatics, 2023

    2023

  15. [23]

    Dsff-gan: A novel stain transfer network for generating immunohistochemical image of endometrial cancer,

    Yihao Ma, Yiqiong Zhang, Zhengrong Wang, Juan Li, Yuehong Miao, Fan Yang, and Wei Pan, “Dsff-gan: A novel stain transfer network for generating immunohistochemical image of endometrial cancer,” Computers in Biology and Medicine, vol. 170, pp. 108046, 2024

    2024

  16. [24]

    Staindiff: Transfer stain styles of his- tology images with denoising diffusion probabilistic models and self- ensemble,

    Yiqing Shen and Jing Ke, “Staindiff: Transfer stain styles of his- tology images with denoising diffusion probabilistic models and self- ensemble,” inInternational Conference on Medical Image Computing and Computer-Assisted Intervention. Springer, 2023, pp. 549–559

    2023

  17. [25]

    Pix2pix-based stain-to-stain translation: A solution for robust stain normalization in histopathology images analysis,

    Pegah Salehi and Abdolah Chalechale, “Pix2pix-based stain-to-stain translation: A solution for robust stain normalization in histopathology images analysis,” in2020 International conference on machine vision and image processing (MVIP). IEEE, 2020, pp. 1–7

    2020

  18. [26]

    Bci: Breast cancer immunohistochemical image generation through pyramid pix2pix,

    Shengjie Liu, Chuang Zhu, Feng Xu, Xinyu Jia, Zhongyue Shi, and Mulan Jin, “Bci: Breast cancer immunohistochemical image generation through pyramid pix2pix,” inProceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2022, pp. 1815–1824

    2022

  19. [27]

    Adaptive supervised patchnce loss for learning h&e-to-ihc stain translation with inconsistent groundtruth image pairs,

    Fangda Li, Zhiqiang Hu, Wen Chen, and Avinash Kak, “Adaptive supervised patchnce loss for learning h&e-to-ihc stain translation with inconsistent groundtruth image pairs,” inInternational Conference on Medical Image Computing and Computer-Assisted Intervention. Springer, 2023,...

    2023

  20. [28]

    Gans trained by a two time-scale update rule converge to a local nash equilibrium,

    Martin Heusel, Hubert Ramsauer, Thomas Unterthiner, Bernhard Nessler, and Sepp Hochreiter, “Gans trained by a two time-scale update rule converge to a local nash equilibrium,”Advances in neural information processing systems, vol. 30, 2017

    2017

  21. [29]

    Con- trastive learning for unpaired image-to-image translation,

    Taesung Park, Alexei A Efros, Richard Zhang, and Jun-Yan Zhu, “Con- trastive learning for unpaired image-to-image translation,” inComputer Vision–ECCV 2020: 16th European Conference, Glasgow, UK, August 23–28, 2020, Proceedings, Part IX 16. Springer, 2020, pp. 319–345

    2020

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.