Research map / Label-efficient and robust learning / Recognition and long-tail
DGR-MIL: Exploring Diverse Global Representation in Multiple Instance Learning for Whole Slide Image Classification
Wenhui Zhu, Xiwen Chen, Peijie Qiu, Aristeidis Sotiras, Abolfazl Razi, Yalin 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 positive instance alignment module to global vectors (ablation) | DGR-MIL · CAMELYON16 | AUC: 0.922 -> 0.944 | Improved |
| add DPP diversity loss on top of positive alignment (ablation) | DGR-MIL · CAMELYON16 | AUC: 0.944 -> 0.957 | Improved |
| add DPP diversity loss on top of positive alignment (ablation) | DGR-MIL · TCGA-NSCLC | AUC: 0.956 -> 0.963 | Improved |
| add tokenized global vector (ablation) | DGR-MIL · CAMELYON16 | AUC: 0.935 -> 0.957 | Improved |
| add tokenized global vector (ablation) | DGR-MIL · TCGA-NSCLC | AUC: 0.957 -> 0.963 | Improved |
| global vectors via cross-attention only, without learning strategies (ablation) | DGR-MIL · CAMELYON16 / TCGA-NSCLC | AUC: 0.922 and 0.928 baseline | Improved |
| increase number of global vectors K beyond optimum (K=5 CAMELYON16, K=3 TCGA) | DGR-MIL · CAMELYON16 / TCGA-NSCLC | performance | Hurt |
| increase loss balance weights lambda_tri and lambda_div to 0.5 or 1.0 | DGR-MIL · CAMELYON16 | classification performance | Hurt |
| reduce loss balance weights to 0.01 | DGR-MIL · CAMELYON16 | performance | Hurt |
| use DGR-MIL instead of DTFD-MIL (AFS) with ResNet-50 features | MIL aggregation · CAMELYON16 | accuracy: 0.908 -> 0.917 | 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
- Exploring Low-Rank Property in Multiple Instance Learning for Whole Slide Image Classification (ICLR 2023 · similarity 0.92)
- Dual-Stream Multiple Instance Learning Network for Whole Slide Image Classification With Self-Supervised Contrastive Learning (CVPR 2021 · similarity 0.91)
- cDP-MIL: Robust Multiple Instance Learning via Cascaded Dirichlet Process (ECCV 2024 · similarity 0.91)
- Dynamic Policy-Driven Adaptive Multi-Instance Learning for Whole Slide Image Classification (CVPR 2024 · similarity 0.90)
- CaMIL: Causal Multiple Instance Learning for Whole Slide Image Classification (AAAI 2024 · similarity 0.90)
- ViLa-MIL: Dual-scale Vision-Language Multiple Instance Learning for Whole Slide Image Classification (CVPR 2024 · similarity 0.89)
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.
