AI Paper Atlas

Research map / Label-efficient and robust learning / Recognition and long-tail

Flow-MIL: Constructing Highly-expressive Latent Feature Space For Whole Slide Image Classification Using Normalizing Flow

Yingfan Ma, Bohan An, Ao Shen, Mingzhi Yuan, Minghong Duan, Manning Wang

ICCV 2025 · Label-efficient and robust learning › Recognition and long-tail · approach: Loss design and classifier theory · 1 citations (OpenAlex)

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 changedWhereMeasuredEffect
add reconstruction loss Lrec (inverse flow reconstruction) (ablation)Flow-MIL · CAMELYON16bag-level AUC: 0.9378 -> 0.9437Improved
add latent alignment loss Llatent (ablation)Flow-MIL · CAMELYON16bag-level AUC: 0.9025 -> 0.9437Improved
add LSP guidance for instance pseudo-labels (ablation)Flow-MIL · CAMELYON16instance-level AUC: 0.9189 -> 0.9277Improved
add LSP guidance for instance pseudo-labels (ablation)Flow-MIL · CAMELYON16bag-level AUC: 0.8987 -> 0.9437Improved
add reconstruction loss Lrec (ablation)Flow-MIL · CAMELYON16instance-level AUC: 0.9246 -> 0.9277Improved
use k=5 GMM components instead of 2 (ablation)Flow-MIL · CAMELYON16bag-level AUC: 0.9256 -> 0.9437Improved
use k=5 GMM components instead of 6 (ablation)Flow-MIL · CAMELYON16bag-level AUC: 0.9066 -> 0.9437Improved
vary k on TCGA-NSCLC (ablation)Flow-MIL · TCGA-NSCLCbag-level AUC: 0.9806-0.9859No clear effect
map features into LSES via normalizing flow (overlap of KDE curves)Flow-MIL · CAMELYON16KDE overlap area: 0.729 -> 0.568Improved
use Flow-MIL instead of fully supervised on bag-levelFlow-MIL · CAMELYON16bag-level AUC: 0.8621 -> 0.9437Improved
use Flow-MIL instead of TransMIL on NSCLCFlow-MIL · TCGA-NSCLCbag-level AUC: 0.9830 -> 0.9859Improved

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.

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

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy