Youlee Han

I am an undergraduate student majoring in Computer Science & Engineering at Ewha Womans University, currently working as an intern at PAI Lab.

My research interests include 3D vision, particularly human-object interaction and 3D scene understanding. My previous research focused on active learning for semantic segmentation and continual learning in multimodal settings.

Email / CV / Scholar / Github

Publication

* indicates equal contribution

Diffusion-Driven Two-Stage Active Learning for Low-Budget Semantic Segmentation
Jeongin Kim, Wonho Bae, YouLee Han, Giyeong Oh, Youngjae Yu, Danica J. Sutherland, Junhyug Noh
NeurIPS, 2025
arXiv / Code

We propose a two-stage active learning framework for semantic segmentation under extreme budget constraints. Our method leverages pre-trained diffusion models for representative candidate selection and introduces a novel uncertainty score to pinpoint the most informative pixels for annotation.

CrossBack: Selective Knowledge Transfer for Cross-modal Devices in Federated Continual Learning
Minju Cho*, Jisoo Jeong*, YouLee Han*
Korean Artificial Intelligence Association Summer Conference, 2025
PDF / Code

We propose CrossBack, a decentralized federated continual learning framework for heterogeneous edge devices that enables selective knowledge transfer across different modalities and input scales without relying on a central server.

This website is based on Jon Barron’s template .