👋 About Me
Hello! I’m Dabin Kim, a Postdoctoral Researcher in the Data Science and Artificial Intelligence Lab (DSAIL) at the Department of Electrical and Computer Engineering, Seoul National University, working with Prof. Seong-Ro Yoon. I am also affiliated with Ulsan National Institute of Science and Technology (UNIST) through the InnoCORE Fellow Program.
I received my Ph.D. in Mechanical and Aerospace Engineering from Seoul National University in 2026, advised by Prof. H. Jin Kim at the Lab for Autonomous Robotics (LARR). I also received my B.S. in Mechanical and Aerospace Engineering from Seoul National University, graduating summa cum laude.
My research focuses on safe and generalizable robot autonomy, with current interests in robot safety, robot manipulation, vision-language-action (VLA) models, and safety for generalist robot policies. During my Ph.D., I built a background in sensor-based planning, visual SLAM, safety-critical control, and safe motion planning.
📰 News
- Feb 2026 - One paper is accepted to ICRA 2026.
- Jan 2026 - Two papers are accepted to RA-L.
- Dec 2025 - I succesfully defended my Ph.D.!
- Oct 2025 - I am honored to have received the Top Prize in the Korea Aerospace Industries (KAI) Aerospace Paper Award.
- Sep 2025 - Our paper was published to IEEE Access.
- May 2025 - I’ll be presenting at ICRA 2025 in Atlanta for paper presentation and participating in the Doctoral Consortium!
- Mar 2025 - 🔍 Post for our recent paper released: [Safety-Critical Control for Visual SLAM]
- Jan 2025 - Our paper was accepted to ICRA 2025.
- Dec 2024 - Gave an oral presentation at the CDC 2024.
- Oct 2024 - Our survey paper was accepted to IJCAS (International Journal of Control, Automation, and Systems).
- Sep 2024 - Visited the University of Padova and gave a talk on perception-aware planning and control.
- Jul 2024 - Our paper was accepted to CDC 2024.
📚 Selected Publications
HERE: Hierarchical Active Exploration of Radiance Field with Epistemic Uncertainty Minimization
Taekbeom Lee*, Dabin Kim*, Youngseok Jang, and H. Jin Kim
2026 RA-L
Active 3D scene reconstruction framework based on neural radiance fields for high-fidelity mapping.
Enhancing Feature Tracking Reliability for Visual Navigation using Real-Time Safety Filter
Dabin Kim*, Inkyu Jang*, Youngsoo Han, Sunwoo Hwang, and H. Jin Kim
2025 ICRA
A real-time safety filter enhances visual feature tracking by modifying robot velocity using a quadratic program.
Estimation of Constraint Admissible Invariant Set with Neural Lyapunov Function
Dabin Kim and H. Jin Kim
2024 CDC
Learning-based estimation of CAPI sets for general reference tracking and real-time constrained control using neural Lyapunov functions.