Pith. sign in

REVIEW 3 major objections 5 minor 28 references

PTCMIL: Multiple Instance Learning via Prompt Token Clustering for Whole Slide Image Analysis

T0 review · 3 major / 5 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read A Vision Transformer with learnable prompt tokens as cluster centers outperforms prior MIL methods on eight whole-slide image datasets, in both classification and survival analysis.

desk verdict Solid new MIL aggregator with a novel prompt-token clustering mechanism, but the SOTA claim needs statistical support. read the letter →

arxiv 2507.18848 v1 pith:EMUUYEHK submitted 2025-07-24 cs.CV cs.AI

Beidi Zhao , SangMook Kim , Hao Chen , Chen Zhou , Zu-hua Gao , Gang Wang , Xiaoxiao Li This is my paper
classification cs.CVcs.AI
keywords multipleinstancelearningwholeslideimageanalysisprompttokenclusteringVisionTransformercomputationalpathologysurvivalprototypepoolingweaklysupervised
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

This paper proposes PTCMIL, a multiple-instance-learning method that reads a whole slide image—a giga-pixel digitized biopsy—as a bag of patch tokens and compresses it into a small set of cluster prototypes. Its central claim is that clustering and slide-level prediction should be optimized together, unlike earlier prototype-based MIL methods that first cluster patches unsupervised and then train a predictor on the prototypes. The method introduces learnable prompt tokens as cluster centers inside a Vision Transformer, assigns each patch to a prompt by inner-product affinity, and lets the classification or survival loss reshape those assignments. The authors report that this end-to-end design outperforms existing MIL aggregators on eight whole-slide datasets and produces clusters that correspond to interpretable tissue types such as tumor, stroma, alveoli, and vessels.

What carries the argument

The load-bearing object is a set of $C$ learnable prompt tokens, one per cluster, that enter a global Transformer alongside the patch tokens and the class token. After one self-attention pass, each patch token is assigned to a cluster by a softmax over inner products with the prompt tokens (Eq. (2)); the assignment is differentiable, so the downstream classification or survival loss reshapes the clusters. An orthogonality regularizer $L_{\text{reg}} = \|\bar{P}^T \bar{P} - I\|^2$ plus a moving-average update keeps the prompt tokens from collapsing, then a shared local Transformer per cluster and weighted centroid merging compress each cluster into one prototype, and the prototypes are pooled with the class token for prediction. This stack is what lets the paper claim that clustering and prediction are optimized jointly rather than in two stages.

What would settle it

Re-run the Camelyon16 and TCGA-NSCLC experiments with the prompt tokens frozen at their orthonormal initial values so that gradients never reshape the cluster assignments; if accuracy and AUC do not drop materially, end-to-end clustering is not the cause of the improvements.

Watch

Extended reading notes

Core claim

On the paper's own terms, the discovery is that clustering and prediction need not be decoupled in whole-slide image analysis: a small number of trainable prompt vectors inside a Vision Transformer can serve as cluster centroids, the assignment of patches to centroids is differentiable through a softmax inner product, and the task loss therefore shapes the clusters. The paper reports that this end-to-end coupling is also the efficient option, because it avoids computing all pairwise patch similarities over tens of thousands of tokens, and that it yields the highest accuracy, AUC, kappa, and concordance index among the compared methods on the eight datasets tested, for example a Camelyon16 AUC of 99.60 with UNI features and a PANDA kappa of 0.937. It further reports that the resulting clusters line up with recognizable tissue types and that a few-shot transfer variant is more accurate than baseline adaptations.

Load-bearing premise

The whole argument rests on the hope that ordinary similarity scores between patch tokens and prompt tokens, after one pass through the Transformer, divide a slide's tissue into stable and meaningful groups; if those groups collapse or drift, the prototypes become uninformative and the reported gains vanish.

Editorial extensions

If this is right

  • The two-stage design of prototype-based MIL—unsupervised clustering first, supervised prediction second—is not necessary; a single ViT can cluster and predict under one loss.
  • A model trained on one cancer type can be adapted to another with about twenty labeled slides by finetuning only the classifier and prototypes, so the approach suits low-data clinical settings.
  • The same prototype representation serves classification and survival analysis, meaning one pretrained aggregator can be redirected to time-to-event tasks without architectural changes.
  • Learned clusters correspond to interpretable tissue types, so slide-level predictions come with a spatial map of where each cluster fires, which is useful for pathology review.

Reading between the lines

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

  • An untested stress test: replace the raw inner-product affinity in Eq. (2) with a learned projection and temperature; if the gains survive, end-to-end prototype supervision, not the specific prompt geometry, is the active ingredient.
  • The prompts could serve as a stain or institution harmonization layer: because they are few, shared, and task-trained, finetuning them on a small local cohort might adapt a slide model across scanners without touching the feature extractor, which goes beyond the paper's few-shot experiments.
  • The paper's stability across cluster counts of 3 to 9 suggests that cluster number could be chosen automatically, for example by a validation-based search or by prototype redundancy, which the authors list as future work.
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

3 major / 5 minor

Summary. The paper proposes PTCMIL, a multiple instance learning aggregator for whole slide images that inserts learnable prompt tokens into a ViT backbone, treats the prompts as cluster prototypes via a soft inner-product assignment (Eq. 2), merges tokens inside each cluster into weighted prototypes, and pools prototype tokens with a class token for classification and survival tasks. Experiments on eight datasets with CTransPath and UNI features report accuracy, AUC, Cohen's kappa, or c-index against multiple prior MIL methods, together with few-shot adaptation results and ablations of the clustering, merging, and pooling modules.

Significance. If the empirical claims hold, PTCMIL offers a conceptually clean end-to-end alternative to two-stage clustering MIL, with limited added parameters and naturally interpretable cluster maps. The authors provide code, systematic ablations of the three architectural components, and visual evidence that the learned clusters correspond to meaningful tissue types. The main contribution is the unification of task-aligned clustering and prediction, and the paper avoids circular evaluation because the representation is trained end-to-end and assessed on external benchmarks. The significance is currently tempered by gaps in the statistical support for the headline state-of-the-art claim and by an uninformative one-class adaptation evaluation.

major comments (3)
  1. [Sec. 3.3, Table 1, Table 2] The central claim that PTCMIL 'outperforms state-of-the-art methods' is not statistically established as presented. Several leading margins are within one standard deviation of the best baseline or are ties: in Table 2 the LUAD c-index is 0.688 for both PTCMIL and ILRA, and the BRCA c-index of 0.745 vs PANTHER's 0.729 is within PANTHER's reported standard deviation of 0.08; in Table 1 the CTransPath PANDA kappa is 0.928 for both PTCMIL and ILRA. The paper reports no paired significance tests, no confidence intervals for kappa or c-index, and no correction for multiple comparisons across eight datasets and two feature extractors. Please add paired permutation tests or bootstrap confidence intervals for the head-to-head comparisons, and temper the abstract and Sec. 3.3 wording to the level of support the statistics provide.
  2. [Sec. 3.1, Table 1] The in-house prostate adaptation evaluation is uninformative as reported. The dataset is described as '1-class' and contains 749 cancerous slides, and the paper reports only accuracy for this column. On a one-class test set, a constant classifier can achieve high accuracy, so this result does not measure the model's ability to distinguish cancer from normal tissue or to adapt to a new domain. Please evaluate on a class-balanced or multi-class cohort and report sensitivity/specificity, AUC, or balanced accuracy; otherwise, remove the adaptability claim based on this column.
  3. [Sec. 2.1, Eq. (2), Prompt Updating] The load-bearing clustering mechanism relies on a raw inner-product softmax over prompt tokens, with no temperature, learned projection, or normalization, and collapse prevention depends on orthogonal initialization, the soft orthogonality regularizer Lreg, and a moving average update. The paper provides no quantitative evidence that the prompts remain diverse and that the assignments are informative after training. Please report cluster assignment entropy or purity, prompt token orthogonality during training, and sensitivity to the number of clusters C, the regularization weight alpha, and the decay factor theta; also compare against a version with a learned projection or temperature in the affinity. Such an analysis would directly support the claim that the prompt-token clustering geometry is stable and meaningful.
minor comments (5)
  1. [Table 4] The ablation table is very difficult to interpret because the rows are not explicitly labeled and the checkmark/dash structure is unclear. Please give each row a named configuration, e.g., 'without clustering', 'without merging', 'without prototype pooling', so the reader can map rows to the text in Sec. 3.5.
  2. [Table 4] Two accuracy values in Table 4 appear to have formatting errors ('94 1.87' for TCGA-NSCLC and '95 0.87' for TCGA-CRC); please correct the decimal formatting.
  3. [References] References [18] and [19] are listed as the same PANTHER paper with identical titles and venue; please remove the duplicate and renumber.
  4. [Sec. 2.2] The text says 'the cluster index vector a = ... for i in [C]' and later refers to 'cluster i' and 'for cluster i' where the index should be c; please fix the variable naming for clarity.
  5. [Sec. 2.3, Eq. (6)] The censoring indicator cj is defined as 1 for patients who live longer than the follow-up period and 0 for patients who pass away, which is the opposite of the common convention. Please state the convention explicitly in the text to avoid confusion with the survival likelihood formula.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: PTCMIL's end-to-end prompt-token clustering is trained against held-out benchmarks; no prediction reduces to a fitted input or to a self-citation chain.

full rationale

The derivation chain in Sec. 2 is a differentiable architecture: Eq. (2) computes a soft assignment from inner products between patch tokens and learnable prompt tokens; Eq. (3) is a soft-orthogonality regularizer; Eq. (4) runs local transformers per cluster; prototypes are weighted averages of cluster token features; Eq. (5) is the classification loss plus the regularizer. Nothing is defined in terms of the final test metric. The prompt tokens and prototype weights are trained on labeled training folds and evaluated on external standard benchmarks (Camelyon16, TCGA, PANDA, TCGA survival), so the reported AUC, kappa, accuracy, and c-index are not fitted quantities renamed as predictions. The paper's central novelty claim (end-to-end alignment of clustering with the downstream objective) is exactly what the training procedure implements; this is a design property, not a circular derivation. The only self-references are citations to baseline methods or background clustering work (e.g., ref. [24] includes co-authors Hao Chen and Xiaoxiao Li; refs. [20,25] include Hao Chen), but these are used as comparisons or prior methods, not as the evidence establishing PTCMIL's performance, so they are not load-bearing. The statistical fragility noted by a skeptic (overlaps within one standard deviation, no paired tests) is a correctness and evidence-strength issue and is explicitly outside the circularity definition. Accordingly, no circular step can be quoted and exhibited, and the honest score is 0.

Assumptions & free parameters 4 free parameters · 5 assumptions · 0 invented entities

The central claim rests on standard MIL assumptions, frozen feature extractors, and two ad hoc geometric assumptions about prompt-token affinities and collapse prevention. There are no new physical entities. The main free parameters are the cluster count and regularization weights, both chosen per dataset.

free parameters (4)
  • Number of clusters C = 7 for Camelyon16; 5 for TCGA and PANDA
    Hyperparameter set per dataset; Fig 3 shows sensitivity, with a peak at 5 for TCGA-NSCLC and CRC. No nested validation is described.
  • Regularization weight alpha = 0.1 for Camelyon16 and PANDA; 0.2 for TCGA datasets
    Chosen by hand; weights the orthogonality penalty in Eq (5).
  • Prompt decay factor theta = 0.9
    Controls moving average of prompt updates; chosen without sensitivity analysis.
  • Optimizer settings = learning rate 2e-4, weight decay 1e-5
    Adam settings from Sec 3.2; standard but tuned.
assumptions (5)
  • domain assumption Slide-level label is a function of the patch instances (standard MIL premise)
    The whole pipeline assumes that a bag label can be predicted from aggregated patch instances; stated implicitly in Sec 1 and used throughout.
  • domain assumption CTransPath and UNI features capture morphology needed for clustering and prediction
    Feature extractors are frozen (Sec 3.2); all clustering and prediction quality depends on these representations.
  • ad hoc to paper Raw inner product between patch and prompt tokens yields a useful soft assignment
    Eq (2) uses inner products as softmax logits without temperature, projection, or normalization; the paper gives no geometric justification for this choice.
  • domain assumption Discrete-time proportional hazards model holds for the learned representation
    Sec 2.3 uses a Cox-style discrete hazard with linear predictor beta Hfinal following [7]; if proportional hazards fails, survival c-index results are harder to interpret.
  • ad hoc to paper Orthogonal initialization and soft penalty prevent prompt collapse
    Prompt initialization and Lreg are heuristic; no guarantee that orthogonality is preserved after training.

how reviews work

0 comments
Cite this review

Pith. "Pith review of PTCMIL: Multiple Instance Learning via Prompt Token Clustering for Whole Slide Image Analysis." pith.science (2026). https://pith.science/paper/EMUUYEHK

@misc{pith2026250718848,
  author       = {Pith},
  title        = {Pith review of: PTCMIL: Multiple Instance Learning via Prompt Token Clustering for Whole Slide Image Analysis},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/EMUUYEHK}},
  note         = {Machine review of arXiv:2507.18848}
}
read the original abstract

Multiple Instance Learning (MIL) has advanced WSI analysis but struggles with the complexity and heterogeneity of WSIs. Existing MIL methods face challenges in aggregating diverse patch information into robust WSI representations. While ViTs and clustering-based approaches show promise, they are computationally intensive and fail to capture task-specific and slide-specific variability. To address these limitations, we propose PTCMIL, a novel Prompt Token Clustering-based ViT for MIL aggregation. By introducing learnable prompt tokens into the ViT backbone, PTCMIL unifies clustering and prediction tasks in an end-to-end manner. It dynamically aligns clustering with downstream tasks, using projection-based clustering tailored to each WSI, reducing complexity while preserving patch heterogeneity. Through token merging and prototype-based pooling, PTCMIL efficiently captures task-relevant patterns. Extensive experiments on eight datasets demonstrate its superior performance in classification and survival analysis tasks, outperforming state-of-the-art methods. Systematic ablation studies confirm its robustness and strong interpretability. The code is released at https://github.com/ubc-tea/PTCMIL.

Figures

Figures reproduced from arXiv: 2507.18848 by the authors.

Figure 1
Figure 1. Overview of PTCMIL. (a) Overall framework, with patch feature tokens, [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Visualization and interpretation of PTCMIL. (a) Clustering maps, cluster [PITH_FULL_IMAGE:figures/full_fig_p008_2.png] view at source ↗
Figure 3
Figure 3. Variation in the number of clusters [PITH_FULL_IMAGE:figures/full_fig_p009_3.png] view at source ↗

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

28 extracted references · 17 canonical work pages

  1. [23]

    In: The Eleventh International Conference on Learning Representations (2023)

    Xiang, J., Zhang, J.: Exploring low-rank property in multiple instance learning for whole slide image classification. In: The Eleventh International Conference on Learning Representations (2023)

    2023

  2. [20]

    Iteratively Coupled Multiple Instance Learning from Instance to Bag Classifier for Whole Slide Image Classification

    Wang, H., Luo, L., Wang, F., Tong, R., Chen, Y.W., Hu, H., Lin, L., Chen, H.: Iteratively coupled multiple instance learning from instance to bag classifier for whole slide image classification. arXiv preprint arXiv:2303.15749 (2023) PTCMIL 11

    work page Pith review arXiv 2023

  3. [18]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

    Song, A.H., Chen, R.J., Ding, T., Williamson, D.F., Jaume, G., Mahmood, F.: Morphological prototyping for unsupervised slide representation learning in com- putational pathology. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 11566–11578 (2024)

    work page 2024

  4. [19]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)

    Song, A.H., Chen, R.J., Ding, T., Williamson, D.F., Jaume, G., Mahmood, F.: Morphological prototyping for unsupervised slide representation learning in com- putational pathology. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). pp. 11566–11578 (June 2024)

    work page 2024

  5. [1]

    Jama318(22), 2199–2210 (2017)

    Bejnordi, B.E., Veta, M., Van Diest, P.J., Van Ginneken, B., Karssemeijer, N., Litjens, G., Van Der Laak, J.A., Hermsen, M., Manson, Q.F., Balkenhol, M., et al.: Diagnostic assessment of deep learning algorithms for detection of lymph node metastases in women with breast cancer. Jama318(22), 2199–2210 (2017)

    2017

  6. [2]

    Nature medicine 28(1), 154–163 (2022)

    Bulten, W., Kartasalo, K., Chen, P.H.C., Ström, P., Pinckaers, H., Nagpal, K., Cai, Y., Steiner, D.F., Van Boven, H., Vink, R., et al.: Artificial intelligence for diagnosis and gleason grading of prostate cancer: the panda challenge. Nature medicine 28(1), 154–163 (2022)

    work page 2022

  7. [3]

    Cancer Genome Atlas Research Network, J., et al.: The cancer genome atlas pan- cancer analysis project. Nat. Genet45(10), 1113–1120 (2013)

    work page 2013

  8. [4]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

    Chan, T.H., Cendra, F.J., Ma, L., Yin, G., Yu, L.: Histopathology whole slide image analysis with heterogeneous graph representation learning. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 15661–15670 (2023)

    work page 2023

Show all 28 references
  1. [5]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

    Chen, R.J., Chen, C., Li, Y., Chen, T.Y., Trister, A.D., Krishnan, R.G., Mahmood, F.: Scaling vision transformers to gigapixel images via hierarchical self-supervised learning. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 16144–161...

    2022

  2. [6]

    Nature Medicine30(3), 850–862 (2024)

    Chen, R.J., Ding, T., Lu, M.Y., Williamson, D.F., Jaume, G., Song, A.H., Chen, B., Zhang, A., Shao, D., Shaban, M., et al.: Towards a general-purpose foundation model for computational pathology. Nature Medicine30(3), 850–862 (2024)

    2024

  3. [7]

    In: Proceedings of the IEEE/CVF Inter- national Conference on Computer Vision (ICCV)

    Chen, R.J., Lu, M.Y., Weng, W.H., Chen, T.Y., Williamson, D.F., Manz, T., Shady, M., Mahmood, F.: Multimodal co-attention transformer for survival pre- diction in gigapixel whole slide images. In: Proceedings of the IEEE/CVF Inter- national Conference on Computer Vision (ICCV)...

    2021

  4. [8]

    In: Proceedings of the thirteenth international conference on ar- tificial intelligence and statistics

    Glorot, X., Bengio, Y.: Understanding the difficulty of training deep feedforward neural networks. In: Proceedings of the thirteenth international conference on ar- tificial intelligence and statistics. pp. 249–256. JMLR Workshop and Conference Proceedings (2010)

    2010

  5. [9]

    In: Proceedings of the AAAI conference on artificial intelligence

    Hou, W., Yu, L., Lin, C., Huang, H., Yu, R., Qin, J., Wang, L.: Hˆ 2-mil: explor- ing hierarchical representation with heterogeneous multiple instance learning for whole slide image analysis. In: Proceedings of the AAAI conference on artificial intelligence. vol. 36, pp. 933–9...

    2022

  6. [10]

    In: International conference on machine learning

    Ilse,M.,Tomczak,J.,Welling,M.:Attention-baseddeepmultipleinstancelearning. In: International conference on machine learning. pp. 2127–2136. PMLR (2018)

    2018

  7. [11]

    Prentice-Hall, Inc

    Jain, A.K., Dubes, R.C.: Algorithms for clustering data. Prentice-Hall, Inc. (1988)

    1988

  8. [12]

    In: European Conference on Computer Vision

    Jia, M., Tang, L., Chen, B.C., Cardie, C., Belongie, S., Hariharan, B., Lim, S.N.: Visual prompt tuning. In: European Conference on Computer Vision. pp. 709–727. Springer (2022)

    2022

  9. [13]

    In: Pro- ceedings of the IEEE/CVF conference on computer vision and pattern recognition

    Li, B., Li, Y., Eliceiri, K.W.: Dual-stream multiple instance learning network for whole slide image classification with self-supervised contrastive learning. In: Pro- ceedings of the IEEE/CVF conference on computer vision and pattern recognition. pp. 14318–14328 (2021)

    2021

  10. [14]

    Nature biomedical engineering5(6), 555–570 (2021)

    Lu, M.Y., Williamson, D.F., Chen, T.Y., Chen, R.J., Barbieri, M., Mahmood, F.: Data-efficient and weakly supervised computational pathology on whole-slide images. Nature biomedical engineering5(6), 555–570 (2021)

    2021

  11. [15]

    In: Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision

    Marin, D., Chang, J.H.R., Ranjan, A., Prabhu, A., Rastegari, M., Tuzel, O.: To- ken pooling in vision transformers for image classification. In: Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision. pp. 12–21 (2023)

    2023

  12. [16]

    Advances in neural infor- mation processing systems34, 13937–13949 (2021)

    Rao, Y., Zhao, W., Liu, B., Lu, J., Zhou, J., Hsieh, C.J.: Dynamicvit: Efficient vision transformers with dynamic token sparsification. Advances in neural infor- mation processing systems34, 13937–13949 (2021)

    2021

  13. [17]

    Advances in neural information processing systems34, 2136–2147 (2021)

    Shao, Z., Bian, H., Chen, Y., Wang, Y., Zhang, J., Ji, X., et al.: Transmil: Trans- former based correlated multiple instance learning for whole slide image classifica- tion. Advances in neural information processing systems34, 2136–2147 (2021)

    2021

  14. [21]

    Meta-Radiology p

    Wang, J., Liu, Z., Zhao, L., Wu, Z., Ma, C., Yu, S., Dai, H., Yang, Q., Liu, Y., Zhang, S., et al.: Review of large vision models and visual prompt engineering. Meta-Radiology p. 100047 (2023)

    2023

  15. [22]

    Medical image analysis81, 102559 (2022)

    Wang, X., Yang, S., Zhang, J., Wang, M., Zhang, J., Yang, W., Huang, J., Han, X.: Transformer-based unsupervised contrastive learning for histopathological image classification. Medical image analysis81, 102559 (2022)

    2022

  16. [24]

    Computerized Medical Imaging and Graphics 97, 102053 (2022)

    Yan, J., Chen, H., Li, X., Yao, J.: Deep contrastive learning based tissue clustering for annotation-free histopathology image analysis. Computerized Medical Imaging and Graphics 97, 102053 (2022)

    2022

  17. [25]

    In: International Conference on Medical Image Computing and Computer-Assisted Intervention

    Yang, S., Wang, Y., Chen, H.: Mambamil: Enhancing long sequence modeling with sequence reordering in computational pathology. In: International Conference on Medical Image Computing and Computer-Assisted Intervention. pp. 296–306. Springer (2024)

    2024

  18. [26]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

    Zeng, W., Jin, S., Liu, W., Qian, C., Luo, P., Ouyang, W., Wang, X.: Not all tokens are equal: Human-centric visual analysis via token clustering transformer. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 11101–11111 (2022)

    2022

  19. [27]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

    Zhang, H., Meng, Y., Zhao, Y., Qiao, Y., Yang, X., Coupland, S.E., Zheng, Y.: Dtfd-mil: Double-tier feature distillation multiple instance learning for histopathol- ogy whole slide image classification. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern ...

    2022

  20. [28]

    In: European Conference on Computer Vision

    Zhu, W., Chen, X., Qiu, P., Sotiras, A., Razi, A., Wang, Y.: Dgr-mil: Exploring diverse global representation in multiple instance learning for whole slide image classification. In: European Conference on Computer Vision. pp. 333–351. Springer (2024)

    2024

Pith tools

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