AI Paper Atlas

Research map / Label-efficient and robust learning / Semi-supervised learning

Towards Generic Semi-Supervised Framework for Volumetric Medical Image Segmentation

Haonan Wang, Xiaomeng Li

NeurIPS 2023 · Label-efficient and robust learning › Semi-supervised learning · approach: Semi-supervised medical segmentation · 16 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 diffusion encoder/decoder (ablation)A&D · MMWHS MR to CT (UDA)Dice: 79.2 -> 90.1Improved
add diffusion encoder/decoder (ablation)A&D · Synapse 20%Dice: 56.7 -> 60.9Improved
add SVDA augmentation (ablation)A&D · MMWHS MR to CT (UDA)Dice: 84.6 -> 90.1Improved
add SVDA augmentation (ablation)A&D · Synapse 20%Dice: 55.3 -> 60.9Improved
add difficulty-aware re-weighting (DRS) (ablation)A&D · Synapse 20%Dice: 52.0 -> 60.9Improved
add difficulty-aware re-weighting (DRS) (ablation)A&D · MMWHS MR to CT (UDA)Dice: 85.8 -> 90.1Improved
add Reparameterize & Smooth pseudo-label strategy (ablation)A&D · MMWHS MR to CT (UDA)Dice: 87.0 -> 90.1Improved
add Reparameterize & Smooth pseudo-label strategy (ablation)A&D · Synapse 20%Dice: 58.7 -> 60.9Improved
use shared diffusion encoder with decoupled decoders instead of three separate networks (ablation)A&D · MMWHS MR to CT (UDA)Dice: 76.14 -> 90.14Improved
use pure diffusion network instead of A&D architecture (ablation)A&D · MMWHS MR to CT (UDA)Dice: 90.14 -> 45.63Hurt
use pure V-Net instead of A&D architecture (ablation)A&D · MMWHS MR to CT (UDA)Dice: 90.14 -> 82.19Hurt
increase diffusion time step from 100 to 1000 (ablation)A&D · MMWHS CT to MRDice: 69.7 -> 71.4Improved

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