# Learning from Partial Label Proportions for Whole Slide Image Segmentation

arXiv: https://arxiv.org/abs/2405.09041 · HTML: https://mlsanity.com/paper/2405.09041/

## Key facts

- Cited by 6 papers in the map (data through 14 Oct 2026).
- Builds on 14 of its 27 references that are in the map.
- First posted on arXiv 15 May 2024.

## Builds on (most cited first, 14 of 14)

- [TransMIL: Transformer based Correlated Multiple Instance Learning for Whole Slide Image Classification](https://mlsanity.com/paper/2106.00908/) (2021)
- [Dual-stream Multiple Instance Learning Network for Whole Slide Image Classification with Self-supervised Contrastive Learning](https://mlsanity.com/paper/2011.08939/) (2020)
- [Revisiting Multiple Instance Neural Networks](https://mlsanity.com/paper/1610.02501/) (2016)
- [Multi-scale Domain-adversarial Multiple-instance CNN for Cancer Subtype Classification with Unannotated Histopathological Images](https://mlsanity.com/paper/2001.01599/) (2020)
- [Additive MIL: Intrinsically Interpretable Multiple Instance Learning for Pathology](https://mlsanity.com/paper/2206.01794/) (2022)
- [Bi-directional Weakly Supervised Knowledge Distillation for Whole Slide Image Classification](https://mlsanity.com/paper/2210.03664/) (2022)
- [Learning from Label Proportions with Consistency Regularization](https://mlsanity.com/paper/1910.13188/) (2019)
- [Deep multi-class learning from label proportions](https://mlsanity.com/paper/1905.12909/) (2019)
- [Co-training for Demographic Classification Using Deep Learning from Label Proportions](https://mlsanity.com/paper/1709.04108/) (2017)
- [Learning from Label Proportions with Generative Adversarial Networks](https://mlsanity.com/paper/1909.02180/) (2019)
- [Learning Whole-Slide Segmentation from Inexact and Incomplete Labels using Tissue Graphs](https://mlsanity.com/paper/2103.03129/) (2021)
- [Learning from Label Proportion with Online Pseudo-Label Decision by Regret Minimization](https://mlsanity.com/paper/2302.08947/) (2023)
- [MixBag: Bag-Level Data Augmentation for Learning from Label Proportions](https://mlsanity.com/paper/2308.08822/) (2023)
- [Proportion Estimation by Masked Learning from Label Proportion](https://mlsanity.com/paper/2405.04815/) (2024)

## Built on by (most cited first, 6 of 6)

- [Theoretical Proportion Label Perturbation for Learning from Label Proportions in Large Bags](https://mlsanity.com/paper/2408.14130/) (2024)
- [Weakly-Supervised Domain Adaptation with Proportion-Constrained Pseudo-Labeling](https://mlsanity.com/paper/2506.22301/) (2025)
- [Learning from Similarity Proportion Loss for Classifying Skeletal Muscle Recovery Stages](https://mlsanity.com/paper/2505.04150/) (2025)
- [Investigating ECG Diagnosis with Ambiguous Labels using Partial Label Learning](https://mlsanity.com/paper/2512.11095/) (2025)
- [Leveraging Label Proportion Prior for Class-Imbalanced Semi-Supervised Learning](https://mlsanity.com/paper/2603.02957/) (2026)
- [Semantic Segmentation for Histopathology using Learned Regularization based on Global Proportions](https://mlsanity.com/paper/2604.24347/) (2026)

## Abstract (arXiv)

In this paper, we address the segmentation of tumor subtypes in whole slide images (WSI) by utilizing incomplete label proportions. Specifically, we utilize `partial' label proportions, which give the proportions among tumor subtypes but do not give the proportion between tumor and non-tumor. Partial label proportions are recorded as the standard diagnostic information by pathologists, and we, therefore, want to use them for realizing the segmentation model that can classify each WSI patch into one of the tumor subtypes or non-tumor. We call this problem ``learning from partial label proportions (LPLP)'' and formulate the problem as a weakly supervised learning problem. Then, we propose an efficient algorithm for this challenging problem by decomposing it into two weakly supervised learning subproblems: multiple instance learning (MIL) and learning from label proportions (LLP). These subproblems are optimized efficiently in the end-to-end manner. The effectiveness of our algorithm is demonstrated through experiments conducted on two WSI datasets.