mlsanityData to 7 Oct 2026

From Correlation to Causation: Max-Pooling-Based Multi-Instance Learning Leads to More Robust Whole Slide Image Classification

arXiv 2408.09449 ↗First posted 18 Aug 2024cs.CV cs.LGcited by 7

Not on the map. Why?

Abstract (arXiv)

In whole slide images (WSIs) analysis, attention-based multi-instance learning (MIL) models are susceptible to spurious correlations and degrade under domain shift. These methods may assign high attention weights to non-tumor regions, such as staining biases or artifacts, leading to unreliable tumor region localization. In this paper, we revisit max-pooling-based MIL methods from a causal perspective. Under mild assumptions, our theoretical results demonstrate that max-pooling encourages the model to focus on causal factors while ignoring bias-related factors. Furthermore, we discover that existing max-pooling-based methods may overfit the training set through rote memorization of instance features and fail to learn meaningful patterns. To address these issues, we propose FocusMIL, which couples max-pooling with an instance-level variational information bottleneck (VIB) to learn compact, predictive latent representations, and employs a multi-bag mini-batch scheme to stabilize optimization. We conduct comprehensive experiments on three real-world datasets and one semi-synthetic dataset. The results show that, by capturing causal factors, FocusMIL exhibits significant advantages in out-of-distribution scenarios and instance-level tumor region localization tasks.

Key facts. Cited by 7 papers in the map (data through 7 Oct 2026). Builds on 27 of its 55 references that are in the map. First posted on arXiv 18 Aug 2024.

Builds on

27 of 55 references.

References · 27

  1. 2017Attention Is All You Needcited by 59,932
  2. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  3. 2018Attention-based Deep Multiple Instance Learningcited by 1,077
  4. 2016Deep Variational Information Bottleneckcited by 915
  5. 2015Deep Learning and the Information Bottleneck Principlecited by 793
  6. 2020Data Efficient and Weakly Supervised Computational Pathology on Whole Slide Imagescited by 738
  7. 2021TransMIL: Transformer based Correlated Multiple Instance Learning for Whole Slide Image Classificationcited by 549
  8. 2020Dual-stream Multiple Instance Learning Network for Whole Slide Image Classification with Self-supervised Contrastive Learningcited by 473
  9. 2021Self-Supervised Learning with Data Augmentations Provably Isolates Content from Stylecited by 308
  10. 2022DTFD-MIL: Double-Tier Feature Distillation Multiple Instance Learning for Histopathology Whole Slide Image Classificationcited by 254
  11. 2020Whole Slide Images based Cancer Survival Prediction using Attention Guided Deep Multiple Instance Learning Networkscited by 178
  12. 2021Whole Slide Images are 2D Point Clouds: Context-Aware Survival Prediction using Patch-based Graph Convolutional Networkscited by 124
15 more
  1. 2016Revisiting Multiple Instance Neural Networkscited by 119
  2. 2023Interventional Bag Multi-Instance Learning On Whole-Slide Pathological Imagescited by 91
  3. 2024Feature Re-Embedding: Towards Foundation Model-Level Performance in Computational Pathologycited by 66
  4. 2023Attention-Challenging Multiple Instance Learning for Whole Slide Image Classificationcited by 61
  5. 2022Additive MIL: Intrinsically Interpretable Multiple Instance Learning for Pathologycited by 58
  6. 2023Multiple Instance Learning Framework with Masked Hard Instance Mining for Whole Slide Image Classificationcited by 58
  7. 2023Task-specific Fine-tuning via Variational Information Bottleneck for Weakly-supervised Pathology Whole Slide Image Classificationcited by 54
  8. 2022Bi-directional Weakly Supervised Knowledge Distillation for Whole Slide Image Classificationcited by 42
  9. 2022Benchmarking the Robustness of Deep Neural Networks to Common Corruptions in Digital Pathologycited by 26
  10. 2023Inherently Interpretable Time Series Classification via Multiple Instance Learningcited by 23
  11. 2023Rethinking Multiple Instance Learning for Whole Slide Image Classification: A Good Instance Classifier is All You Needcited by 21
  12. 2022Multi-Instance Causal Representation Learning for Instance Label Prediction and Out-of-Distribution Generalizationcited by 17
  13. 2024Memorization in deep learning: A surveycited by 14
  14. 2023Reproducibility in Multiple Instance Learning: A Case For Algorithmic Unit Testscited by 11
  15. 2024AEM: Attention Entropy Maximization for Multiple Instance Learning based Whole Slide Image Classificationcited by 10

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 1; 2026: 6.

7 papers.

  1. 2025VistaWise: Building Cost-Effective Agent with Cross-Modal Knowledge Graph for Minecraftcited by 13
  2. 2026VideoStir: Understanding Long Videos via Spatio-Temporally Structured and Intent-Aware RAGcited by 5
  3. 2026TRACE: Trajectory Recovery for Continuous Mechanism Evolution in Causal Representation Learningcited by 4
  4. 2026Multi-View Stenosis Classification Leveraging Transformer-Based Multiple-Instance Learning Using Real-World Clinical Datacited by 1
  5. 2026The Weakest Link Tells It All: Outcome-Supervised Process Reward Modeling via Learnable Credit Assignmentcited by 0
  6. 2026From Patches to Evidence Balls: Class-Conditioned Evidence Retrieval for Few-Shot Whole Slide Image Classificationcited by 0
  7. 2026Weakly-supervised Kidney Tumor Classification from CT Scans with Multi-Instance Learning and Anatomical Filteringcited by 0