Research map / Label-efficient and robust learning / Noisy labels
Norma: A Noise Robust Memory-Augmented Framework for Whole Slide Image Classification
Yu Bai, Bo Zhang, Zheng Zhang, Shuo Yan, Zibo Ma, Wu Liu, Xiuzhuang Zhou, Xiangyang Gong, Wendong Wang
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 a global memory sink of full size (ablation) | Norma · CAMELYON-16 | AUC: 83.2 -> 95.1 | Improved |
| add a global memory sink of full size (ablation) | Norma · TCGA-LUNG | AUC: 93.5 -> 96.9 | Improved |
| use full memory sink rather than half (ablation) | Norma · CAMELYON-16 | AUC: 93.6 -> 95.1 | Improved |
| use RPEG instead of absolute positional encoding (ablation) | Norma · CAMELYON-16 | AUC: 92.3 -> 95.1 | Improved |
| use RPEG instead of relative positional encoding (ablation) | Norma · CAMELYON-16 | AUC: 93.9 -> 95.1 | Improved |
| keep spatial order of patches rather than random sequence (ablation) | Norma · CAMELYON-16 | AUC: 90.1 -> 95.1 | Improved |
| add positional encoding (RPEG) (ablation) | Norma · CAMELYON-16 | AUC: 89.7 -> 95.1 | Improved |
| use two training cycles instead of one (ablation) | Norma · CAMELYON-16 | AUC: 92.3 -> 95.1 | Improved |
| increase cycles beyond two (ablation) | Norma · CAMELYON-16 | AUC: 95.1 -> 94.7 (3 cycles) | No clear effect |
| add noise detection loss Lnoise (ablation) | Norma · CAMELYON-16 | AUC: 93.9 -> 95.1 | Improved |
| add supervised contrastive loss Lctr (ablation) | Norma · CAMELYON-16 | AUC: 94.6 -> 95.1 | Improved |
| use learnable soft pseudo labels instead of hard argmax labels for noise detection (ablation) | Norma · three WSI datasets | AUC | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- TransMIL — TransMIL: Transformer based Correlated Multiple Instance Learning for Whole Slide Image Classification (NeurIPS 2021) · 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
- Boosting Multiple Instance Learning Models for Whole Slide Image Classification: A Model-Agnostic Framework Based on Counterfactual Inference (AAAI 2024 · similarity 0.87)
- FAST: A Dual-tier Few-Shot Learning Paradigm for Whole Slide Image Classification (NeurIPS 2024 · similarity 0.86)
- Queryable Prototype Multiple Instance Learning with Vision-Language Models for Incremental Whole Slide Image Classification (AAAI 2025 · similarity 0.86)
- OODML: Whole Slide Image Classification Meets Online Pseudo-Supervision and Dynamic Mutual Learning (AAAI 2025 · similarity 0.85)
- ViLa-MIL: Dual-scale Vision-Language Multiple Instance Learning for Whole Slide Image Classification (CVPR 2024 · similarity 0.84)
- Learning from Memory: Non-Parametric Memory Augmented Self-Supervised Learning of Visual Features (ICML 2024 · similarity 0.84)
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.
