CV

Curriculum Vitae of Yujin Bae.

Contact Information

Name Yujin Bae
Professional Title M.S. Student in Robotics, KAIST
Email yujinbae@kaist.ac.kr
Location KAIST, 291 Daehak-ro, Yuseong-gu, Daejeon 34141, Republic of Korea
Advisor Prof. Kuk-Jin Yoon (Visual Intelligence Lab)
Website https://jinyubae.github.io

Education

  • 2025 - Present

    M.S.
    Korea Advanced Institute of Science and Technology (KAIST)
    Robotics Program
    • Advisor: Prof. Kuk-Jin Yoon (Visual Intelligence Lab)
  • 2019 - 2025

    B.S.
    Ewha Womans University
    Software Science and Engineering (Cyber Security Major)
    • GPA: 4.17 / 4.3
    • Ranked 1st in major
  • 2023 - 2023

    Exchange Student
    University of Vermont
    Computer and Information Sciences
    • Funded by Mirae Asset Overseas Exchange Scholarship

Honors and Awards

  • 2023
    Mirae Asset Overseas Exchange Scholarship
    Mirae Asset Park Hyeon Joo Foundation

    28th cohort

  • 2022
    Grand Prize (Minister's Award) — Blockchain Challenge
    Ministry of Science and ICT, Republic of Korea

    National competition

  • 2019 – 2023
    Honors Scholarship
    Ewha Womans University

    Top of major (Spring 2020)
    Top 2% of class (Fall 2019)
    Top 6% of class (Fall 2020, Spring 2022, Spring 2023)

Research Interests

Anticipation: forecasting human motion and action semantics from egocentric video with multimodal LLMs
Understanding: efficient egocentric video understanding guided by human behavior
Transfer: learning dexterous manipulation policies from human video

Research Projects

  • 2026 - Present

    Full-Body Motion Prediction from Sparse Observations via Expert Distillation
    Visual Intelligence Lab, KAIST
    • Developing a lightweight model that forecasts future motion from sparse observations via on-policy distillation, jointly decoding action intent as text to disambiguate multimodal futures
  • 2025 - 2026

    Development of Core Technologies for Autonomous Driving in Unstructured Off-Road Environments
    Funded by Hanwha Aerospace
    • Contributed to a moving object tracking algorithm that fuses LiDAR, Radar, and RGB sensors for robust perception in unstructured off-road environments
    • Collected robot driving datasets (ROS bag) across off-road and diverse operational settings, and annotated 2D bounding box tracks of moving objects

Research Experience

  • 2024

    Research Intern (Full-time)
    HCI Tech Lab, Graduate School of Culture Technology, KAIST
    Advisor: Prof. Sang Ho Yoon
    • Contributed to a hand pressure estimation algorithm augmenting forearm sEMG with 3D hand posture, surveying existing 3D hand pose estimation methods and integrating them into the end-to-end pipeline
  • 2022

    Research Intern (Full-time)
    Humanoid Generalization Lab, Graduate School of AI, KAIST
    Advisor: Prof. Beomjoon Kim

Languages

Korean : Native
English : Fluent (TOEFL 5.5)

Skills

Programming Languages: Python, C++
Frameworks: PyTorch, ROS