Hyomuk Kim
I am currently a graduate student pursuing an M.S. in Electrical and Computer Engineering at the University of California, San Diego (UCSD), specializing in Intelligent Systems, Robotics, and Control (EC80).
Prior to joining UCSD, I was a Staff Engineer and Roboticist at the Robot Center, Samsung Research. During my tenure there, I worked on mobile robot navigation in collaboration with Mid-Eum Choi and Junghyun Kwon, diving deep into Visual SLAM and Navigation systems, specifically focusing on 3D localization, mapping, and robust path planning.
My research experience at Samsung also extends to AI applications; I conducted projects on neural networks for Brain-Machine Interfaces advised by Inchul Hwang, and developed a Neural Speech Synthesis engine advised by Chanwoo Kim at the Global AI Center.
Earlier in my career, I served as a hardware engineer for 5 years at the Visual Display Division of Samsung Electronics, developing circuit systems for various TV models. Additionally, I served as a planner and manager for a year at C-Lab, leading a 6-member team in developing a user content archive platform utilizing multi-device interfaces.
This website is dedicated to sharing my research journey. Please refer to my Curriculum Vitae for more details.
Education
- University of California, San Diego (UCSD), La Jolla, CA
- Master of Science in Electrical and Computer Engineering (Intelligent Systems, Robotics, and Control)
Sep 2025 – Present - Relevant Coursework: Statistical Learning, Random Processes, Linear Algebra and application, Linear Systems Theory, Sensing & Estimation in Robotics
- Master of Science in Electrical and Computer Engineering (Intelligent Systems, Robotics, and Control)
- Chung-Ang University, Seoul, Korea
- Bachelor of Science in Electrical & Electronics Engineering
Mar 2007 – Feb 2011 - ROTC (Reserve Officers’ Training Corps) Cadet, 49th, Republic of Korea Army
Mar 2009 – Feb 2011
- Bachelor of Science in Electrical & Electronics Engineering
Professional Experiences
- Samsung Research, Seoul, Korea
- Staff Engineer (Robotics), Robot Intelligence Team
Apr. 2021 – Aug. 2024 - Engineer (Deep Learning), Global AI Center
Jan. 2020 – Apr. 2021 - Engineer (Deep Learning), Language & Voice Team
Apr. 2018 – Jan. 2020
- Staff Engineer (Robotics), Robot Intelligence Team
- Samsung Electronics, Suwon, Korea
- Project Leader, C-Lab (Visual Display Div.)
Jun. 2014 – Jun. 2015 - Engineer (Circuitry), TV R&D Lab (Visual Display Div.)
Apr. 2013 – Apr. 2018
- Project Leader, C-Lab (Visual Display Div.)
Military Service
- Republic of Korea Army, Korea
- First Lieutenant, Reconnaissance Battalion (60th Div.)
Mar. 2011 – Jun. 2013
- First Lieutenant, Reconnaissance Battalion (60th Div.)
Research Interests
Robotics, AI / ML, Robot Learning, Computer Vision, Autonomous Driving / Navigation
Research Objectives
General Objectives
Explore robots' ability to understand and navigate diverse environments safely both indoors and outdoors when performing a variety of tasks, also considering learning-based methodologies for continuous intelligence improvement.
Specific Objectives
Assess autonomous robots and vehicles' ability to navigate complex 3D environments with various static and dynamic obstacles using limited sensor data like images or scans, while considering practical constraints.
Develop techniques to factorize scene elements from visual sensors, incorporating semantic objects and diverse features, utilizing both novel neural networks and classical tracking methods. Investigate geometric parameterization for detailed 3D space representation for future planning. Explore methods for integrating multiple visual and sensory inputs dynamically for adaptability across varied sensor configurations to minimize geometric ambiguity.
Discuss how robots enhance navigation through diverse learning approaches (such as deep reinforcement learning, imitation learning, neural SLAM, and even meta-learning), and describe their incremental expansion of semantic perception and mapping, through which emphasize the importance of evolving robot navigation through continuous learning.
Evaluate robots' adaptability in unique environments like underwater, forests, and outer space, assessing their seamless navigation capabilities across these environments to achieve long-lasting navigation for versatile robots.