Download the complete CV (PDF)
Education
- Ph.D. Candidate in Mechanical Engineering (Mar. 2023 - Feb. 2028, Expected)
- Korea Advanced Institute of Science and Technology (KAIST), Daejeon, Korea
- Advisor: Kuk-Jin Yoon
- GPA: 3.81/4.3
- M.S. in Mechanical Engineering (Sep. 2021 - Feb. 2023)
- Korea Advanced Institute of Science and Technology (KAIST), Daejeon, Korea
- Advisor: Kuk-Jin Yoon
- GPA: 4.10/4.3
- B.S. in Mechanical Engineering and Electrical Engineering (Double Major) (Mar. 2015 - Feb. 2021)
- Korea Advanced Institute of Science and Technology (KAIST), Daejeon, Korea
- GPA: 3.76/4.3, cum laude
Research Interests
- Foundation Model Optimization and Adaptation
- Multi-task, multi-objective, and non-convex optimization
- Parameter-efficient adaptation, model merging, and knowledge distillation (including on-policy distillation)
- Test-time adaptation and in-context learning
- Embodied Intelligence
- Vision-language-action models and uncertainty-guided reasoning
- Robust perception and motion planning for autonomous systems
Research Experience
- Project Leader (Mar. 2025 - Present)
- Core Technology Development for Autonomous Driving in Unstructured Off-Road Environments
- Funded by Hanwha Aerospace. Real-world deployment target in unstructured off-road terrain.
- Project Leader (Mar. 2024 - Mar. 2025)
- Unmanned Swarm CPS Research Lab
- Funded by Agency for Defense Development (ADD). Multi-agent perception and coordination for autonomous robotic systems.
- Project Leader (Mar. 2022 - Feb. 2024)
- Development of a 2kg-Class Small Electro-Optical Tracking System for Drones
- Funded by Institute of Civil-Military Technology Cooperation. Multi-modal (EO/IR) detection and tracking deployed on drone hardware.
- Project Member (Mar. 2021 - Feb. 2022)
- Unmanned Swarm CPS Research Lab
- Funded by Agency for Defense Development (ADD). Multi-agent perception research for autonomous robotic systems.
Patents
- Apparatus and Method for Learning Multiple Tasks in a Partial Labeling Environment
- Republic of Korea Patent Application No. 10-2026-0089936 (Filed May 18, 2026)
- Jae Young Lee, Seung Uk Ahn, Young Woo Seo, Kuk-Jin Yoon, Wooseong Jeong, Je Gyeong Cho, Young Ho Yoon
- Apparatus and Method for Generating Image Dataset for Learning and Evaluation of Image Stitching
- Republic of Korea Patent No. 10-2645640 (Registered Mar. 5, 2024)
- Kuk-Jin Yoon, Hyeokjun Kweon, Yoonsu Kang, Wooseong Jeong, Hyeonseong Kim, Hwi-Sung Park
- Apparatus and Method for Image Stitching
- Republic of Korea Patent No. 10-2571530 (Registered Aug. 23, 2023)
- Kuk-Jin Yoon, Hyeokjun Kweon, Yoonsu Kang, Wooseong Jeong, Hyeonseong Kim, Hwi-Sung Park
Honors and Awards
- Qualcomm Innovation Fellowship Korea 2025 – Finalist (2025)
- Selected paper: Interaction-Merged Motion Planning: Effectively Leveraging Diverse Motion Datasets for Robust Planning
Academic Service
- NeurIPS (2024, 2025)
- ICLR (2025, 2026)
- CVPR (2025, 2026)
- ICML (2025)
- ICCV (2025)
- ECCV (2026)
- BMVC (2026)
- AAAI (2027)
- IJCV (2026)
- IEEE TMM (2026)
Skills
- Languages: Python (proficient), C++ (proficient), MATLAB
- ML Frameworks: PyTorch, JAX, NumPy, OpenCV
- Efficient Training: PyTorch DDP, DeepSpeed, flash-attention, xformers
- Hugging Face: Transformers, Diffusers, PEFT, Accelerate
- Experiment & DevOps: Weights & Biases, TensorBoard, Docker, Git
- Robotics: ROS, ROS2
- Spoken Languages: Korean (native), English (fluent)