Thanks for visiting my website! โœจ
I am an incoming CS PhD student at UNC-Chapel Hill, MURGe-Lab (Advisor: Prof.Mohit Bansal).
Before starting PhD, I was advised by Prof.Jinkyu Kim at Korea University and Prof.Sung Ju Hwang at KAIST.
I am interested in Robust / Interpretable AI and pursuing to build applicable models to make contributions to the real world!
In particular, I want to make user-interactive AI which can help underrepresented people all around the world.

My current research interests include

  • User-Interactive Multimodal
  • Out-of-Distribution Generalization
  • Continual Learning

๐Ÿ”ฅ News

  • 2024.05: ย ๐ŸŽ‰ One paper got accepted to ICMLโ€™24 ย 
  • 2024.04: ย ๐ŸŽ‰ One paper got accepted to CVPRWโ€™24 ย 
  • 2024.03: ย ๐ŸŽ“ I received BE degree at KU and will start PhD journey at the University of North Carolina at Chapel Hill MURGe-Lab!
  • 2023.03: ย ๐ŸŽ‰ I started my new research journey at KAIST MLAI LAB (Under the supervision of Prof. Sung Ju Hwang) ย 
  • 2023.01: ย ๐ŸŽ‰ One paper is oral-presented at WACVโ€™23 ย 
  • 2022.12: ย ๐ŸŽ‰ I finished my first research internship at NAVER LABS well ๐Ÿš™ ย 
  • 2022.10: ย ๐ŸŽ‰ One paper is presented at ECCVโ€™22 ย 

๐Ÿ’– Publications

ICML 2024
sym
  • BECoTTA: Input-dependent Online Blending of Experts for Continual Test-time Adaptation [paper] [project] [code]
    • Daeun Lee*, Jaehong Yoon*, Sung Ju Hwang (* denotes equal contribution)
    • This work was done at the KAIST MLAI Lab.
CVPRW 2024
sym
  • HD Maps are Lane Detection Generalizers: A Novel Generative Framework for Single-Source Domain Generalization
    • Daeun Lee, Minhyeok Heo, Jiwon Kim
    • This work was done at NAVER LABS.
WACV 2023
sym
  • Resolving Class Imbalance for LiDAR-based Object Detector by Dynamic Weight Average and Contextual Ground Truth Sampling [paper]
    • Daeun Lee, Jinkyu Kim
    • This work was done at the Korea University VAI Lab (collaborated with Hyundai Motor).
ECCV 2022
sym
  • Bridging the Domain Gap towards Generalization in Automatic Colorization [paper]

    • Hyejin Lee, Daehee Kim, Daeun Lee, Jinkyu Kim and Jaekoo Lee
    • This work was done at the Korea University VAI Lab.
Preprint
sym
  • Trajectory Prediction by Clustering Human Interactions at Multiple Scales

    • Chiho Choi*, Daeun Lee*, Srikanth Malla, Sangjae Bae, Jinkyu Kim.
    • This work was done at the Korea University VAI Lab (collaborated with Honda Research).

๐Ÿ“– Educations

  • 2019.03 - 2024.02, B.E. in Statistics. Korea University. (GPA: 4.0/4.5)

    • 2020 - 2024, Special Scholarship for outstanding students.

๐Ÿ’ป Research Experiences

  • CVPR 2022, ECCV 2022, CVPR 2024, ECCV 2024 Reviewer

  • 2024.03 - current, Contract Researcher, MLAI Lab, South Korea.

  • 2023.03 - 2024.02, Research Intern, MLAI Lab, South Korea.
    • Under the supervision of Prof. Sung Ju Hwang
  • 2022.07 - 2022.12, Research Intern, NAVER LABS, South Korea.
    • Autonomous Driving Perception team
  • 2021.06 - 2022.12, Research Intern, VAI Lab, South Korea.
    • Under the supervision of Prof. Jinkyu Kim
  • 2021.01 - 2021.02, Data Analyst Intern, Mirae Asset Securities, South Korea.

๐ŸŽ– Honors and Awards


๐Ÿซ Teaching Assistant

  • 2023.03 - current, Computer Vision Mentor, 5th/6th NAVER BoostCamp
  • 2022.06 - 2022.10, Computer Vision Mentor, YearDream School

๐Ÿ’ Acknowledgement

I always believe my achievements and learning are from my great advisers and collaboration.
I really appreciate Jinkyu Kim, Jaehong Yoon, Minhyeok Heo, Sung Ju Hwang, Sungbin Lim.