Research map / Label-efficient and robust learning / Semi-supervised learning
Semi-supervised Segmentation of Histopathology Images with Noise-Aware Topological Consistency
Meilong Xu, Xiaoling Hu, Saumya Gupta, Shahira Abousamra, Chao Chen
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add signal topology consistency loss only (ablation) | TopoSemiSeg on CRAG 20% · CRAG (20% labeled) | Betti Matching Error: 12.125 -> 10.150 | Improved |
| add both signal consistency and noisy topology removal losses (ablation) | TopoSemiSeg on CRAG 20% · CRAG (20% labeled) | Betti Matching Error: 12.125 -> 8.575 | Improved |
| decompose persistence diagrams into signal and noise (phi 0.7) instead of direct matching (phi 0) (ablation) | TopoSemiSeg · CRAG (20% labeled) | Betti Matching Error: 14.025 -> 8.575 | Improved |
| retain noise and remove signal instead of the proposed decomposition (ablation) | TopoSemiSeg · CRAG (20% labeled) | Betti Matching Error: 8.575 -> 21.325 | Hurt |
| vary persistence threshold phi from 0.5 to 0.9 (ablation) | TopoSemiSeg · CRAG (20% labeled) | Betti Matching Error: 8.575 to 10.150 | No clear effect |
| add topology loss to PSPNet backbone (ablation) | PSPNet · CRAG (20% labeled) | Betti Matching Error: 70.625 -> 44.150 | Improved |
| EMA decay 0.9999 or 0.99 instead of 0.999 (ablation) | TopoSemiSeg · CRAG (20% labeled) | Betti Matching Error: 8.575 -> 11.850 | Hurt |
| increase topology loss weight lambda_U2 to 0.01 (ablation) | TopoSemiSeg · CRAG (20% labeled) | Dice_Obj: 0.898 -> 0.873 | Hurt |
| add supervised plus unsupervised topology losses on 100% labeled data (ablation) | Fully-supervised UNet++ · CRAG (100% labeled) | Dice_Obj: 0.928 -> 0.913 | Hurt |
| topology consistency on 20% labels vs supervised TopoLoss | TopoSemiSeg · CRAG (20% labeled) | Betti Matching Error: 19.925 -> 8.575 | Improved |
| pixel-wise accuracy vs XNet/URPC on some settings | TopoSemiSeg · MoNuSeg (20% labeled) | Accuracy: 0.909 (URPC) vs 0.908 | No clear effect |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- XNet — XNet: Wavelet-Based Low and High Frequency Fusion Networks for Fully- and Semi-Supervised Semantic Segmentation of Biomedical Images (ICCV 2023) · better on 8 of 8 shared results
- FixMatch — FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidence (NeurIPS 2020) · better on 1 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- MATCH: Multi-faceted Adaptive Topo-Consistency for Semi-Supervised Histopathology Segmentation (NeurIPS 2025 · similarity 0.93)
- Topograph: An Efficient Graph-Based Framework for Strictly Topology Preserving Image Segmentation (ICLR 2025 · similarity 0.88)
- Weakly Supervised Gland Segmentation with Class Semantic Consistency and Purified Labels Filtration (AAAI 2025 · similarity 0.88)
- Prototype-Based Image Prompting for Weakly Supervised Histopathological Image Segmentation (CVPR 2025 · similarity 0.87)
- Dynamic Entity-Masked Graph Diffusion Model for Histopathology Image Representation Learning (AAAI 2025 · similarity 0.86)
- Dynamic Graph Representation with Knowledge-aware Attention for Histopathology Whole Slide Image Analysis (CVPR 2024 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
