Pith. sign in

REVIEW 4 major objections 6 minor 48 references

Equivariant Imaging Biomarkers for Robust Unsupervised Segmentation of Histopathology

T0 review · 4 major / 6 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read Rotation-proof symmetric convolution kernels keep unsupervised histopathology segmentation consistent when images are rotated.

desk verdict A clean empirical demonstration that a custom equivariant kernel improves rotation-consistency of unsupervised histopathology clustering, but the novelty claim is overstated and the pathologist-alignment result is weaker than it looks. read the letter →

arxiv 2505.05689 v1 pith:SKJVG7VG submitted 2025-05-08 eess.IV cs.CVcs.LG

Fuyao Chen , Yuexi Du , Tal Zeevi , Nicha C. Dvornek , John A. Onofrey This is my paper · ORCID
classification eess.IVcs.CVcs.LG
keywords HistopathologyEquivariantCNNSRE-ConvUnsupervisedsegmentationProstatecancerK-meansclusteringRotationrobustnessDigitalpathology
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 tries to establish that a convolutional kernel with built-in rotational symmetry makes unsupervised segmentation of histopathology images stable under arbitrary rotation, without any labeled training data for the target tissue. The authors build a network (SRENet) by replacing every convolution in a ResNet18 with a symmetric rotation-equivariant kernel (SRE-Conv), pre-train it on a public colon-tissue dataset, then cluster its features with K-means to segment prostate tissue-microarray images. Across twelve 30-degree rotations, the same pixels keep their cluster labels far more consistently for SRENet (intra-subject ICC 0.92, Kappa 0.90, Dice 0.90) than for a standard ResNet or the E2CNN rotation-equivariant baseline, and the unsupervised embeddings align closely with pathologist Gleason grade segmentations (Dice 0.91). Why this matters: histopathology images have no meaningful orientation, so a feature extractor that ignores rotation could yield more reliable automated grading and longitudinal monitoring of biopsies.

What carries the argument

The load-bearing object is the symmetric rotation-equivariant convolution (SRE-Conv) kernel: a convolution kernel whose weights are shared along concentric rings around the center, so that rotating the input and then convolving equals convolving and then rotating the output. In this paper it is instantiated in SRENet, a fully convolutional network that replaces every ResNet18 convolution with SRE-Conv layers of kernel sizes [9,9,5,5] across the four stages, followed by equivariant pooling and a 1x1 convolution, with global adaptive pooling at the classifier. The kernel's ring parameterization gives the network local rotational/reflection invariance through a Hadamard product and global equivariance under convolution, with an appended proof of the equivariance property. The unsupervised segmentation pipeline then feeds layer-4 features (scaled to 128x128, tissue-masked) into K-means clustering with K=3, using 2000 sampled features per image for intra-subject analysis and 12,500 aggregated features for inter-subject analysis.

What would settle it

Rotate the test images by 15 degrees, which is not a multiple of the 30-degree grid used in the paper, and measure whether SRENet's intra-subject ICC stays near 0.92; if the value falls to the ResNet baseline level, the equivariance is only true on the tested grid rather than the claimed arbitrary rotation.

Watch

Extended reading notes

Core claim

On its own terms, the paper claims that rotation equivariance in a convolutional feature extractor is sufficient to make unsupervised segmentation consistent under rotation. The mechanism is the centrally symmetric SRE-Conv kernel, whose weights are constant along each ring around the kernel center; because convolution with such a kernel commutes with rotation of the input, the extracted feature maps transform predictably. K-means clustering on layer-4 features then labels each pixel with a tissue-cluster identity, and those labels survive image rotation: SRENet reports intra-subject ICC/Kappa/Dice of 0.92/0.90/0.90 versus 0.86/0.80/0.81 for E2CNN and 0.85/0.82/0.82 for ResNet, with inter-subject metrics similarly highest (0.91/0.90/0.91), all differences significant at p<0.05. A secondary claim is that the unsupervised SRENet embeddings carry pathology-relevant structure: when pathologist Gleason Grade labels are mapped into the embedding space, SRENet achieves Dice 0.91 against 0.82 for E2CNN and 0.83 for ResNet.

Load-bearing premise

The load-bearing premise is that features from a network pre-trained on colon tissue (NCT-CRC) remain informative enough for prostate TMA tissue that K-means clusters reflect biological tissue type; if domain shift breaks this transfer, the reported rotation consistency could instead be measuring stain or background artifacts.

Editorial extensions

If this is right

  • If the central claim holds, a single SRENet pre-trained on one tissue type can produce rotation-consistent unsupervised segmentations on an entirely different tissue type, since the reported inter-subject results were obtained on unseen prostate images after colon-only pre-training.
  • Longitudinal active-surveillance workflows become feasible: repeated prostate biopsies from the same patient could be segmented with the same K-means model and compared directly across time, because pixel-level labels do not scramble when the slide is rotated.
  • Unsupervised equivariant embeddings could serve as a biomarker-discovery tool: the Dice 0.91 agreement with pathologist Gleason grading suggests the clusters capture diagnostically meaningful tissue structure, not just low-level texture.
  • Because the consistency gain holds across K=2, 3, and 4 and with Gaussian mixture clustering in the ablation, the benefit is attributable to the equivariant features rather than to one specific clustering choice.

Reading between the lines

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

  • A testable extension is continuous-angle robustness: the paper evaluates rotations at 30-degree steps, while the equivariance proof is stated for arbitrary angles, so checking 15-degree or 45-degree rotations would show whether the real-world stability matches the theoretical equivariance.
  • If the kernel symmetry transfers, the approach may reduce or replace geometric data augmentation in histopathology self-supervised pre-training, since the network no longer needs to see rotations and flips of the same image to learn orientation-invariant representations.
  • The domain-shift caveat cuts both ways: if a prostate-pretrained SRENet or a multi-organ foundation model preserves the reported consistency, the same biomarker pipeline could plausibly extend to breast, colorectal, and skin histology without re-engineering, but the present paper only demonstrates this on one target organ.
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 / 6 minor

Summary. This paper proposes an unsupervised segmentation pipeline for histopathology based on symmetric rotation-equivariant convolution (SRE-Conv), which is used to build a ResNet18-based feature extractor (SRENet). The method extracts feature maps from a network pre-trained on the NCT-CRC colon dataset, performs K-means clustering on the features, and maps cluster labels back to pixels. The validation uses 50 prostate TMA images from the Gleason 2019 Challenge and measures rotation robustness by fitting K-means on a 0-degree image and applying it to 12 rotated versions, reporting ICC, Cohen's Kappa, and Dice. SRENet is compared with a conventional ResNet and an E2CNN baseline, and the paper claims significantly higher consistency metrics for SRENet, as well as high Dice (0.91) when aligning features with pathologist Gleason grade annotations.

Significance. If the central claims hold, the paper addresses a real and important problem: histopathology images lack a canonical orientation, and deep learning models that are robust to arbitrary rotations would be valuable for clinical deployment. The idea of using equivariant convolutions in an unsupervised segmentation setting is timely, and the comparison with E2CNN is scientifically interesting because both models are nominally equivariant, so the reported advantage of SRENet may reflect differences in how discrete rotations are handled. The paper also ships a public code repository and reports the ablation over K-means cluster numbers and Gaussian mixture clustering, which are useful. However, the current validation has load-bearing weaknesses: the pathologist-alignment evaluation is in-sample, the discrete-equivariance caveat is not discussed, and the acknowledged domain shift between colon pre-training and prostate evaluation is not mitigated. These issues are fixable within the scope of a revision but currently limit the strength of the biomarker claims.

major comments (4)
  1. [Section 4.5 / Appendix A4] The pathologist-alignment evaluation in Appendix A4 is transductive: a kNN classifier is trained on features from the same 25 test subjects on which it is then evaluated. The reported Dice of 0.91 for SRENet therefore does not establish that the unsupervised clusters correspond to Gleason grade groups; it only shows that the feature space can separate already-annotated pixels after being trained on those same pixels. The authors should either train the kNN on the 25 training subjects (or another held-out split) and evaluate on the 25 test subjects, or explicitly reframe this analysis as an in-sample feature-separability diagnostic rather than as evidence of alignment with pathologist segmentations. As written, the abstract's statement about aligning with pathologist Gleason Grade segmentations overstates what the evidence supports.
  2. [Section 3.1 / Appendix A1] The equivariance proof in Appendix A1 treats rotations as continuous transformations, but the experiments rotate discrete images by 30-degree increments. A discretely sampled radially symmetric kernel with shared ring weights is exactly invariant only under rotations that map the pixel grid onto itself, such as multiples of 90 degrees and reflections; for 30-degree rotations, resampling and interpolation break exact equivariance. The paper should explicitly state that the equivariance is approximate in the discrete setting and should connect this to the observed ICC values (which are high but not 1.0). This caveat is central because the term 'equivariant' in the title and throughout the paper suggests an exact property, while the empirical setting only supports approximate robustness.
  3. [Section 5 (Discussion)] The authors acknowledge the domain shift between NCT-CRC colon pre-training and prostate TMA evaluation, but this limitation is not mitigated. If the features do not transfer meaningfully to prostate tissue, the reported cluster consistency could reflect staining or background artifacts rather than biologically meaningful tissue compartments. Because all models share the same pre-training, the rotation-robustness comparison remains internally valid, but the broader biomarker interpretation for prostate cancer requires either a quantitative evaluation of cluster validity on prostate-specific tissue (for example, using a held-out prostate dataset or an external unsupervised clustering quality metric) or a more tempered claim about clinical utility.
  4. [Section 4.2 (Model Implementation and Baseline Comparisons)] The conventional ResNet baseline is trained without any geometric data augmentation, yet rotation and reflection augmentation is standard practice for histopathology CNNs and would likely improve the robustness of the standard convolution baseline. Without a ResNet trained with such augmentation, the comparison does not reflect a strong conventional baseline, and the claim that equivariant kernels outperform standard convolutions is weakened. The authors mention this as future work; either including this baseline or discussing its absence as a limitation is needed for the comparison to be fair.
minor comments (6)
  1. [Appendix A1] The proof of equivariance is difficult to follow; the change of variables in the integral is not fully justified. A standard proof for radially symmetric kernels, or a citation to a reference, would improve clarity.
  2. [Section 4.2] The number of rotations N used for the E2CNN baseline (e.g., C8 or C12) should be stated, since E2CNN's equivariance is relative to its chosen rotation group. The specific feature-extraction layer used for each model should also be specified, as the comparison may be sensitive to this choice.
  3. [Section 4.3] The definitions of ICC and Dice for cluster labels should be spelled out: the ICC type (one-way vs two-way, single vs average measures), and whether Dice is averaged over all pairwise comparisons of the 12 rotations or computed against the 0-degree segmentation only.
  4. [Section 4.4 / 4.5] Exact p-values for the Wilcoxon rank-sum tests should be reported (or effect sizes with confidence intervals), and the paper should state whether any correction for multiple comparisons across the three metrics and two analyses was applied.
  5. [Abstract / Section 2.1] The phrase 'invariant only to translation' is imprecise: CNNs are translation equivariant, not invariant. Consider rewording to 'equivariant only to translation'.
  6. [Section 5] There is a typo in the first paragraph: 'quantitativel evaluate' should be 'quantitatively evaluate'.

Circularity Check

2 steps flagged · score 6.0 of 10

Pathologist-alignment Dice in App. A4 is transductive (a kNN classifier is trained on labels from the same 25 test images it then scores), and the headline rotation-consistency result is largely a restatement of the radially symmetric kernel design, though comparisons against E2CNN and NCT-CRC classification provide independent content.

  1. fitted input called prediction [Appendix A4 (and summarized in Section 4.5)]
    "define an embedding space ... using a small subset of the unsupervised features (100 sample from each subject) from all subjects in the inter-subject testing cohort; (2) map the ground-truth pathologist labels from this subset onto each point in the embedding space; (3) train a supervised learning classifier (k-nearest neighbor with k=3) within the embedding space; (4) project the feature vectors from all image pixel locations ... and (5) classify the projected features from each image using the trained classifier to segment the image. ..."

    The kNN classifier is fitted to 100 labeled samples taken from every subject in the 25-subject testing cohort, and the Dice is then computed on all pixels of those very same subjects. The classifier has therefore already seen labels from each test image before that image is segmented, so the reported Dice (0.91 for SRENet) measures the supervised probe's fit to the evaluation labels rather than an unsupervised correspondence between SRENet clusters and Gleason grades. No held-out subject or held-out image is used, making the 'alignment to pathologist segmentation' result statistically forced by the training labels.

  2. self definitional [Appendix A1 (kernel design) and Section 4.4 (intra-subject result)]
    "each circular ring from the center represents one trainable parameter. This design shares values among parameters symmetric to the center, providing local rotational and reflection invariance via the Hadamard product and global equivariance under convolution. ... SRENet exhibited higher intra-subject ICC, Kappa and Dice when compared to both E2CNN and ResNet (p<0.05), indicating superior label consistency following rotation."

    The SRE-Conv kernel is radially symmetric by construction: all positions on a circular ring share a single weight. Consequently, when the input image is rotated, the convolution response at the corresponding rotated tissue location is the same feature value, so after unrotating the segmentation the K-means cluster labels are unchanged up to interpolation/resizing artifacts. The intra-subject ICC/Kappa/Dice consistency is therefore a direct consequence of the kernel's definition, not an empirically discovered prediction. The only non-tautological part of this headline result is the comparison against E2CNN, whose group-equivariant feature channels are not scalar-invariant under the same fixed K-means model.

full rationale

The paper has independent content: SRENet is compared with E2CNN, an external state-of-the-art equivariant architecture, and with ResNet; the NCT-CRC pretraining results (Table A3) show empirical accuracy differences (SRENet 95.5 vs E2CNN 93.8 vs ResNet 93.7 on the original test set) that are not forced by the kernel design. However, two load-bearing validations reduce by construction. First, the pathologist-alignment Dice reported in Section 4.5 and Appendix A4 is computed by training a kNN classifier on labeled samples from the very same test cohort that it then segments, so the 0.91 Dice does not establish that SRENet's unsupervised clusters correspond to Gleason grades. Second, the central rotation-robustness claim for SRENet is largely a restatement of the radially symmetric kernel: a kernel whose weights are constant on each circular ring produces rotation-invariant scalar features, so high intra-subject label consistency across rotations is built into the architecture. The paper's honest acknowledgment of domain-shift risk (Section 5) is a limitation, not circularity. Overall, the derivation is partially circular rather than wholly so, because the comparison with E2CNN and the classification pretraining retain independent empirical content.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

The paper introduces no new physical or mathematical entities; SRE-Conv is prior work by the same group. The central claim depends on hand-chosen clustering hyperparameters (K, L, n) and on the transferability of features from colon to prostate tissue, rather than on new fitted constants.

free parameters (3)
  • number of clusters K = 3
    Main results use K=3 K-means clusters. The ablation in Appendix A3 shows K=2 yields even higher consistency metrics, so the headline K=3 is a hand-selected compromise between consistency and tissue-type discrimination.
  • feature extraction layer L = 4
    Features are extracted from the 4th layer of the encoder, scaled to 128x128. No systematic study of other layers is provided, so this is a hand-chosen model component that affects the results.
  • sampled feature count n = 2000 intra-subject, 500 inter-subject per image
    Random subsampling sizes for K-means fitting and assignment are chosen by hand and affect clustering behavior and computational cost.
assumptions (4)
  • standard math Rotation is a measure-preserving isometry and convolution on R^2 is equivariant under rotation: R(h*f) = (R h)*(R f).
    Invoked in Appendix A1 to prove SRE-Conv equivariance; this is a standard result for isometries on Euclidean space.
  • standard math A convolution kernel with constant values on circles centered at the origin satisfies R h = h for rotations and reflections.
    This property of the ring-parameterized SRE kernel is used in the Appendix A1 proof to conclude R(h*f) = h*R(f).
  • domain assumption Features learned from NCT-CRC colon tissue transfer to prostate TMA tissue well enough for K-means to yield biologically meaningful clusters.
    All models are pre-trained on colon data and applied to prostate images. The authors acknowledge this domain shift as a limitation in Section 5, making it a load-bearing premise of the validation.
  • domain assumption Cluster label maps produced by K-means are comparable across rotated versions without explicit label matching.
    The consistency metrics (ICC, Kappa, Dice) assume that equivariant features preserve cluster identities under rotation. This is plausible for exactly equivariant features but is not explicitly verified against label permutations.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Equivariant Imaging Biomarkers for Robust Unsupervised Segmentation of Histopathology." pith.science (2026). https://pith.science/paper/SKJVG7VG

@misc{pith2026250505689,
  author       = {Pith},
  title        = {Pith review of: Equivariant Imaging Biomarkers for Robust Unsupervised Segmentation of Histopathology},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/SKJVG7VG}},
  note         = {Machine review of arXiv:2505.05689}
}
read the original abstract

Histopathology evaluation of tissue specimens through microscopic examination is essential for accurate disease diagnosis and prognosis. However, traditional manual analysis by specially trained pathologists is time-consuming, labor-intensive, cost-inefficient, and prone to inter-rater variability, potentially affecting diagnostic consistency and accuracy. As digital pathology images continue to proliferate, there is a pressing need for automated analysis to address these challenges. Recent advancements in artificial intelligence-based tools such as machine learning (ML) models, have significantly enhanced the precision and efficiency of analyzing histopathological slides. However, despite their impressive performance, ML models are invariant only to translation, lacking invariance to rotation and reflection. This limitation restricts their ability to generalize effectively, particularly in histopathology, where images intrinsically lack meaningful orientation. In this study, we develop robust, equivariant histopathological biomarkers through a novel symmetric convolutional kernel via unsupervised segmentation. The approach is validated using prostate tissue micro-array (TMA) images from 50 patients in the Gleason 2019 Challenge public dataset. The biomarkers extracted through this approach demonstrate enhanced robustness and generalizability against rotation compared to models using standard convolution kernels, holding promise for enhancing the accuracy, consistency, and robustness of ML models in digital pathology. Ultimately, this work aims to improve diagnostic and prognostic capabilities of histopathology beyond prostate cancer through equivariant imaging.

Figures

Figures reproduced from arXiv: 2505.05689 by the authors.

Figure 1
Figure 1. Intra-Subject and Inter-Subject Analysis. We visualize an example image for intra-subject (A) and inter-subject (B) analyses using equivariant learning (SRENet and E2CNN) and standard convolution (ResNet). The TMA image undergoes 30-degree rotation increments (top row). For each rotation angle, the resulting segmentation after unsupervised K-means clustering was plotted and then unrotated back to the original input … view at source ↗
Figure 2
Figure 2. Comparison to Pathologist Segmentation. For 4 TMA image example sub￾jects (column 1), we visualize the pathologist segmentations (column 2) in com￾parison with labeled segmentation maps from our equivariant SRENet model (column 3), rotation equivariant baseline E2CNN (column 4), and conventional non-equivariant baseline ResNet (column 5). Because the K-means cluster fitting was performed independently, the segmentat… view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

48 extracted references · 40 canonical work pages

  1. [1]

    Mitosis domain generalization in histopathology images - the MIDOG challenge

    Marc Aubreville, Nikolas Stathonikos, Christof A Bertram, Robert Klopfleisch, Natalie Ter Hoeve, Francesco Ciompi, Frauke Wilm, Christian Marzahl, Taryn A Donovan, Andreas Maier, Jack Breen, Nishant Ravikumar, Youjin Chung, Jinah Park, Ramin Nateghi, Fattaneh Pourakpour, Rutger H J Fick, Saima Ben Hadj, Mostafa Jahanifar, Adam Shephard, Jakob Dexl, Thomas...

    work page 2023

  2. [2]

    Artificial intelligence for diagnosis and gleason grading of prostate cancer: the PANDA challenge

    Wouter Bulten, Kimmo Kartasalo, Po-Hsuan Cameron Chen, Peter Ström, Hans Pinckaers, Kunal Nagpal, Yuannan Cai, David F Steiner, Hester van Boven, Robert Vink, Christina Hulsbergen-van de Kaa, Jeroen van der Laak, Mahul B Amin, Andrew J Evans, Theodorus van der Kwast, Robert Allan, Peter A Humphrey, Henrik Grönberg, Hemamali Samaratunga, Brett Delahunt, To...

    work page 2022

  3. [3]

    Deep- HiTS : Rotation invariant convolutional neural network for transient detection

    Guillermo Cabrera-Vives, Ignacio Reyes, Francisco Förster, Pablo A Estévez, and Juan-Carlos Maureira. Deep- HiTS : Rotation invariant convolutional neural network for transient detection. ApJ, 836: 0 97, 2017

    work page 2017

  4. [4]

    A program to build E ( N )-equivariant steerable CNNs

    Gabriele Cesa, Leon Lang, and Maurice Weiler. A program to build E ( N )-equivariant steerable CNNs . In International conference on learning representations, 2022

    work page 2022

  5. [5]

    Rotation equivariant and invariant neural networks for microscopy image analysis

    Benjamin Chidester, Tianming Zhou, Minh N Do, and Jian Ma. Rotation equivariant and invariant neural networks for microscopy image analysis. Bioinformatics, 35: 0 i530--i537, 2019

    work page 2019

  6. [6]

    Conditional positional encodings for vision transformers, 2021

    Xiangxiang Chu, Zhi Tian, Bo Zhang, Xinlong Wang, and Chunhua Shen. Conditional positional encodings for vision transformers, 2021. URL https://arxiv.org/abs/2102.10882

    arXiv 2021

  7. [7]

    A coefficient of agreement for nominal scales

    Jacob Cohen. A coefficient of agreement for nominal scales. Educational and Psychological Measurement, 20 0 (1): 0 37–46, April 1960. ISSN 1552-3888. doi:10.1177/001316446002000104. URL http://dx.doi.org/10.1177/001316446002000104

    doi:10.1177/001316446002000104 1960

  8. [8]

    Group equivariant convolutional networks

    Taco Cohen and Max Welling. Group equivariant convolutional networks. In Maria Florina Balcan and Kilian Q Weinberger, editors, Proceedings of The 33rd International Conference on Machine Learning, volume 48 of Proceedings of Machine Learning Research, pages 2990--2999. PMLR, 2016

    work page 2016

Show all 48 references
  1. [9]

    An image is worth 16x16 words: Transformers for image recognition at scale

    Alexey Dosovitskiy, Lucas Beyer, Alexander Kolesnikov, Dirk Weissenborn, Xiaohua Zhai, Thomas Unterthiner, Mostafa Dehghani, Matthias Minderer, Georg Heigold, Sylvain Gelly, Jakob Uszkoreit, and Neil Houlsby. An image is worth 16x16 words: Transformers for image recognition at...

    2020

  2. [10]

    Dvornek, and John A

    Yuexi Du, Jiazhen Zhang, Tal Zeevi, Nicha C. Dvornek, and John A. Onofrey. Sre-conv: Symmetric rotation equivariant convolution for biomedical image classification, 2025. URL https://arxiv.org/abs/2501.09753

    2025 arXiv

  3. [11]

    Use of symmetric kernels for convolutional neural networks

    Viacheslav Dudar and Vladimir Semenov. Use of symmetric kernels for convolutional neural networks. In Recent Developments in Data Science and Intelligent Analysis of Information, pages 3--10. Springer International Publishing, 2019

  4. [12]

    SymNet : Symmetrical filters in convolutional neural networks

    Gregory Dzhezyan and Hubert Cecotti. SymNet : Symmetrical filters in convolutional neural networks. arXiv [cs.CV], 2019

    2019

  5. [13]

    Polar transformer networks

    Carlos Esteves, Christine Allen-Blanchette, Xiaowei Zhou, and Kostas Daniilidis. Polar transformer networks. arXiv [cs.CV], 2017

    2017

  6. [14]

    A rotationally-invariant convolution module by feature map back-rotation

    Patrick Follmann and Tobias Bottger. A rotationally-invariant convolution module by feature map back-rotation. In 2018 IEEE Winter Conference on Applications of Computer Vision (WACV), pages 784--792. IEEE, 2018

    2018

  7. [15]

    Rotated ring, radial and depth wise separable radial convolutions

    Wolfgang Fuhl and Enkelejda Kasneci. Rotated ring, radial and depth wise separable radial convolutions. In 2021 International Joint Conference on Neural Networks (IJCNN), pages 1--8. IEEE, 2021

    2021

  8. [16]

    Rotation equivariant siamese networks for tracking

    D Gupta, Devanshu Arya, and E Gavves. Rotation equivariant siamese networks for tracking. Proc. IEEE Comput. Soc. Conf. Comput. Vis. Pattern Recognit., pages 12357--12366, 2020

    2020

  9. [17]

    Breast cancer multi-classification from histopathological images with structured deep learning model

    Zhongyi Han, Benzheng Wei, Yuanjie Zheng, Yilong Yin, Kejian Li, and Shuo Li. Breast cancer multi-classification from histopathological images with structured deep learning model. Sci. Rep., 7: 0 4172, 2017

    2017

  10. [18]

    RRL :regional rotation layer in convolutional neural networks

    Zongbo Hao, Tao Zhang, Mingwang Chen, and Kaixu Zhou. RRL :regional rotation layer in convolutional neural networks. arXiv [cs.CV], 2022

    2022

  11. [19]

    Deep residual learning for image recognition

    Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. Deep residual learning for image recognition. In 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). IEEE, 2016

    2016

  12. [20]

    Spatial transformer networks

    Max Jaderberg, K Simonyan, Andrew Zisserman, and K Kavukcuoglu. Spatial transformer networks. Adv. Neural Inf. Process. Syst., abs/1506.02025, 2015

    2015 arXiv

  13. [21]

    Black, Larry Goldenberg, and Septimiu E

    Davood Karimi, Guy Nir, Ladan Fazli, Peter C. Black, Larry Goldenberg, and Septimiu E. Salcudean. Deep learning-based gleason grading of prostate cancer from histopathology images—role of multiscale decision aggregation and data augmentation. IEEE Journal of Biomedical and Hea...

    2020

  14. [22]

    Predicting survival from colorectal cancer histology slides using deep learning: A retrospective multicenter study

    Jakob Nikolas Kather, Johannes Krisam, Pornpimol Charoentong, Tom Luedde, Esther Herpel, Cleo-Aron Weis, Timo Gaiser, Alexander Marx, Nektarios A Valous, Dyke Ferber, Lina Jansen, Constantino Carlos Reyes-Aldasoro, Inka Zörnig, Dirk Jäger, Hermann Brenner, Jenny Chang-Claude, ...

    2019

  15. [23]

    CyCNN : A rotation invariant CNN using polar mapping and cylindrical convolution layers

    Jinpyo Kim, Wooekun Jung, Hyungmo Kim, and Jaejin Lee. CyCNN : A rotation invariant CNN using polar mapping and cylindrical convolution layers. arXiv [cs.CV], 2020

    2020

  16. [24]

    Sample efficient semantic segmentation using rotation equivariant convolutional networks

    Jasper Linmans, Jim Winkens, Bastiaan S Veeling, Taco S Cohen, and Max Welling. Sample efficient semantic segmentation using rotation equivariant convolutional networks. arXiv [cs.CV], 2018

    2018

  17. [25]

    Image analysis and machine learning in digital pathology: Challenges and opportunities

    Anant Madabhushi and George Lee. Image analysis and machine learning in digital pathology: Challenges and opportunities. Medical Image Analysis, 33: 0 170--175, 2016

    2016

  18. [26]

    Rotation equivariant vector field networks

    Diego Marcos, M Volpi, N Komodakis, and D Tuia. Rotation equivariant vector field networks. ICCV, pages 5058--5067, 2016

    2016

  19. [27]

    McGraw and S

    Kenneth O. McGraw and S. P. Wong. Forming inferences about some intraclass correlation coefficients. Psychological Methods, 1 0 (1): 0 30--46, 1996. doi:10.1037/1082-989X.1.1.30

    1996 doi

  20. [28]

    Why capsule neural networks do not scale: Challenging the dynamic parse-tree assumption, 2023

    Matthias Mitterreiter, Marcel Koch, Joachim Giesen, and S\" o ren Laue. Why capsule neural networks do not scale: Challenging the dynamic parse-tree assumption, 2023. URL https://arxiv.org/abs/2301.01583

    2023 arXiv

  21. [29]

    RIC - CNN : Rotation-invariant coordinate convolutional neural network

    Hanlin Mo and Guoying Zhao. RIC - CNN : Rotation-invariant coordinate convolutional neural network. Pattern Recognit., 146: 0 109994, 2024

    2024

  22. [30]

    Whole slide imaging versus microscopy for primary diagnosis in surgical pathology: A multicenter blinded randomized noninferiority study of 1992 cases (pivotal study)

    Sanjay Mukhopadhyay, Michael D Feldman, Esther Abels, Raheela Ashfaq, Senda Beltaifa, Nicolas G Cacciabeve, Helen P Cathro, Liang Cheng, Kumarasen Cooper, Glenn E Dickey, Ryan M Gill, Robert P Heaton, Jr, René Kerstens, Guy M Lindberg, Reenu K Malhotra, James W Mandell, Ellen ...

    1992

  23. [31]

    Automatic grading of prostate cancer in digitized histopathology images: Learning from multiple experts

    Guy Nir, Soheil Hor, Davood Karimi, Ladan Fazli, Brian F Skinnider, Peyman Tavassoli, Dmitry Turbin, Carlos F Villamil, Gang Wang, R Storey Wilson, Kenneth A Iczkowski, M Scott Lucia, Peter C Black, Purang Abolmaesumi, S Larry Goldenberg, and Septimiu E Salcudean. Automatic gr...

    2018

  24. [32]

    Skinnider, Peyman Tavassoli, Dmitry Turbin, Carlos F

    Guy Nir, Soheil Hor, Davood Karimi, Ladan Fazli, Brian F. Skinnider, Peyman Tavassoli, Dmitry Turbin, Carlos F. Villamil, Gang Wang, R. Storey Wilson, Kenneth A. Iczkowski, M. Scott Lucia, Peter C. Black, Purang Abolmaesumi, S. Larry Goldenberg, and Septimiu E. Salcudean. Auto...

    2018 doi

  25. [33]

    SPIN : Simplifying polar invariance for neural networks application to vision-based irradiance forecasting

    Quentin Paletta, Anthony Hu, Guillaume Arbod, Philippe Blanc, and Joan Lasenby. SPIN : Simplifying polar invariance for neural networks application to vision-based irradiance forecasting. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pag...

    2022

  26. [34]

    Limitation of capsule networks, 2019

    David Peer, Sebastian Stabinger, and Antonio Rodriguez-Sanchez. Limitation of capsule networks, 2019. URL https://arxiv.org/abs/1905.08744

    2019 arXiv

  27. [35]

    Dynamic routing between capsules, 2017

    Sara Sabour, Nicholas Frosst, and Geoffrey E Hinton. Dynamic routing between capsules, 2017. URL https://arxiv.org/abs/1710.09829

    2017 arXiv

  28. [36]

    Going deeper through the gleason scoring scale: An automatic end-to-end system for histology prostate grading and cribriform pattern detection

    Julio Silva-Rodríguez, Adrián Colomer, María A Sales, Rafael Molina, and Valery Naranjo. Going deeper through the gleason scoring scale: An automatic end-to-end system for histology prostate grading and cribriform pattern detection. Comput. Methods Programs Biomed., 195: 0 105...

    2020

  29. [37]

    Visualizing data using t- SNE

    L Van der Maaten and G Hinton. Visualizing data using t- SNE . J. Mach. Learn. Res., 2008

    2008

  30. [38]

    Predicting breast tumor proliferation from whole-slide images: The TUPAC16 challenge

    Mitko Veta, Yujing J Heng, Nikolas Stathonikos, Babak Ehteshami Bejnordi, Francisco Beca, Thomas Wollmann, Karl Rohr, Manan A Shah, Dayong Wang, Mikael Rousson, Martin Hedlund, David Tellez, Francesco Ciompi, Erwan Zerhouni, David Lanyi, Matheus Viana, Vassili Kovalev, Vitali ...

    2019

  31. [39]

    General e (2)-equivariant steerable cnns

    Maurice Weiler and Gabriele Cesa. General e (2)-equivariant steerable cnns. Adv. Neural Inf. Process. Syst., 32, 2019

    2019

  32. [40]

    Learning steerable filters for rotation equivariant CNNs

    Maurice Weiler, Fred A Hamprecht, and Martin Storath. Learning steerable filters for rotation equivariant CNNs . In 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition. IEEE, 2018

    2018

  33. [41]

    Pan-tumor CAnine cuTaneous cancer histology ( CATCH ) dataset

    Frauke Wilm, Marco Fragoso, Christian Marzahl, Jingna Qiu, Chloé Puget, Laura Diehl, Christof A Bertram, Robert Klopfleisch, Andreas Maier, Katharina Breininger, and Marc Aubreville. Pan-tumor CAnine cuTaneous cancer histology ( CATCH ) dataset. Sci. Data, 9: 0 588, 2022

    2022

  34. [42]

    Harmonic networks: Deep translation and rotation equivariance

    Daniel E Worrall, Stephan J Garbin, Daniyar Turmukhambetov, and Gabriel J Brostow. Harmonic networks: Deep translation and rotation equivariance. In 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). IEEE, 2017

    2017

  35. [43]

    Rotation invariant gabor convolutional neural network for image classification

    Xiaoqin Yao and Tiecheng Song. Rotation invariant gabor convolutional neural network for image classification. Pattern Recognit. Lett., 162: 0 22--30, 2022

    2022

  36. [44]

    Stable and symmetric filter convolutional neural network

    Raymond Yeh, Mark Hasegawa-Johnson, and Mink N Do. Stable and symmetric filter convolutional neural network. In 2016 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), pages 2652--2656. IEEE, 2016

    2016

  37. [45]

    Wide residual networks

    Sergey Zagoruyko and Nikos Komodakis. Wide residual networks. arXiv [cs.CV], 2016

    2016

  38. [46]

    Dvornek, and John A

    Jiazhen Zhang, Yuexi Du, Nicha C. Dvornek, and John A. Onofrey. Improved vessel segmentation with symmetric rotation-equivariant u-net, 2025. URL https://arxiv.org/abs/2501.14592

    2025 arXiv

  39. [47]

    MTCNet : Multi-task collaboration network for rotation-invariance face detection

    Lifang Zhou, Hui Zhao, and Jiaxu Leng. MTCNet : Multi-task collaboration network for rotation-invariance face detection. Pattern Recognit., 124: 0 108425, 2022

    2022

  40. [48]

    Oriented response networks

    Yanzhao Zhou, Qixiang Ye, Qiang Qiu, and Jianbin Jiao. Oriented response networks. Proc. IEEE Comput. Soc. Conf. Comput. Vis. Pattern Recognit., pages 4961--4970, 2017

    2017

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.