Research map / Label-efficient and robust learning / Recognition and long-tail
Cracking Instance Jigsaw Puzzles: An Alternative to Multiple Instance Learning for Whole Slide Image Analysis
Xiwen Chen, Peijie Qiu, Wenhui Zhu, Hao Wang, Huayu Li, Xuanzhao Dong, Xiaotong Sun, Xiaobing Yu, Yalin Wang, Abolfazl Razi, Aristeidis Sotiras
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 shuffling equivalence regularization (ablation) | Ours [Trans.] · CAMELYON16 (Swin-ViT) | Accuracy: 86.51 -> 89.53 | Improved |
| add shuffling equivalence regularization (ablation) | Ours [CNN] · CAMELYON16 (Swin-ViT) | Accuracy: 85.58 -> 88.11 | Improved |
| add shuffling equivalence regularization (ablation) | Ours [Trans.] · TCGA-NSCLC (Swin-ViT) | Accuracy: 90.60 -> 92.32 | Improved |
| add shuffling equivalence regularization (ablation) | Ours [CNN] · TCGA-NSCLC (Swin-ViT) | Accuracy: 90.21 -> 92.51 | Improved |
| add shuffling equivalence regularization (ablation) | Ours [Trans.] · TCGA-NSCLC (Swin-ViT) | AUC: 95.80 -> 96.40 | Improved |
| add shuffling equivalence regularization (ablation) | Ours [Trans.] · CAMELYON16 (CTransPath) | Accuracy: 95.89 -> 96.64 | Improved |
| add shuffling equivalence regularization (ablation) | Ours [CNN] · CAMELYON16 (CTransPath) | Accuracy: 95.21 -> 96.25 | Improved |
| add shuffling equivalence regularization (ablation) | Ours [Trans.] · TCGA-NSCLC (CTransPath) | Accuracy: 94.53 -> 95.20 | Improved |
| add shuffling equivalence regularization (ablation) | Ours [CNN] · TCGA-NSCLC (CTransPath) | Accuracy: 94.43 -> 95.11 | Improved |
| use CNN instead of transformer in the Siamese jigsaw model | Ours · TCGA-LUAD survival (UNI) | C-Index: 0.677 -> 0.683 | Improved |
| use transformer with equivalence regularization instead of TransMIL | Ours [Trans.] · TCGA-LUAD survival (UNI) | C-Index: 0.609 -> 0.677 | Improved |
| use parallelized dual-branch training (batch size 2) instead of batch size 1 | Ours [Trans.] · Time per iteration | time (ms): 7.8 -> 4.6 | 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 2 of 2 shared results
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
- Dynamic Policy-Driven Adaptive Multi-Instance Learning for Whole Slide Image Classification (CVPR 2024 · similarity 0.88)
- Multiple Instance Learning Framework with Masked Hard Instance Mining for Whole Slide Image Classification (ICCV 2023 · similarity 0.88)
- CaMIL: Causal Multiple Instance Learning for Whole Slide Image Classification (AAAI 2024 · similarity 0.87)
- Dynamic Graph Representation with Knowledge-aware Attention for Histopathology Whole Slide Image Analysis (CVPR 2024 · similarity 0.86)
- Generalizable Whole Slide Image Classification with Fine-Grained Visual-Semantic Interaction (CVPR 2024 · similarity 0.85)
- Queryable Prototype Multiple Instance Learning with Vision-Language Models for Incremental Whole Slide Image Classification (AAAI 2025 · similarity 0.85)
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.
