AI Paper Atlas

Research map / Label-efficient and robust learning / Domain adaptation and generalization

Learning to Generalize over Subpartitions for Heterogeneity-Aware Domain Adaptive Nuclei Segmentation

Jianan Fan, Dongnan Liu, Hang Chang, Weidong Cai

IJCV 2024 · Label-efficient and robust learning › Domain adaptation and generalization · approach: Cross-modality medical segmentation adaptation · 14 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 progressive clustering and separation (PCS) (ablation)Ours · BBBC039 -> KumarPQ: 0.5318 -> 0.5527Improved
add nucleus shape and structure preserving module (NSSP) (ablation)Ours · BBBC039 -> KumarPQ: 0.4558 -> 0.5527Improved
add instance-level disentanglement (ID) (ablation)Ours · BBBC039 -> KumarPQ: 0.5083 -> 0.5527Improved
add global-local style consistency (GLSC) (ablation)Ours · BBBC039 -> KumarPQ: 0.5287 -> 0.5527Improved
also apply clustering-based pseudo subdomain labels in Stage II (ablation)Ours · BBBC039 -> KumarPQ: 0.5527 -> 0.5074Hurt
use K=10 latent subdomains instead of K=4 (ablation)Ours · BBBC039 -> KumarPQ: 0.5366 -> 0.5527Improved
use K=10 latent subdomains instead of K=16 (ablation)Ours · BBBC039 -> KumarPQ: 0.5468 -> 0.5527Improved
use K-Means instead of Mean-Shift clustering (ablation)Ours · BBBC039 -> KumarPQ: 0.5194 -> 0.5527Improved
use our Stage I framework instead of DRIT (ablation)Ours · BBBC039 -> KumarPQ: 0.3978 -> 0.5527Improved
use our Stage I framework instead of DRANet (ablation)Ours · BBBC039 -> KumarPQ: 0.4153 -> 0.5527Improved
apply color normalization to Kumar images (ablation)Ours · BBBC039 -> KumarPQHurt
scale loss weights by 0.3 to 3 (ablation)Ours · BBBC039 -> KumarPQNo clear effect

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