Semi-Supervised Learning in Medical Image

mj lim • TEAM LEADER •

Wanzee Cho​

[Scholar]

About team SSL

● Mission: ​
To develop robust semi-supervised segmentation frameworks for limited labeled data scenarios​

● Scope: ​
Medical image segmentation under annotation scarcity with broader vision applicability​

● Goal: ​
To develop practical SSL-based segmentation systems for clinical AI adoption and real-world environments​​​​​

Available internship topics

● Semi-supervised segmentation with limited labels and consistency-based learning​​

● Large-scale dataset curation and preprocessing for visual learning​​

Our research topics

Semi-supervised Image Segmentation with Pseudo-label Refinement​

● Prompt-Guided Reliable Pseudo-Labeling for Semi-Supervised Segmentation​

Semi-supervised Image Segmentation for Domain Generalization​

● Domain-Generalizable Segmentation via Map Correction Module​