Kyuhwan Yeon

Kyuhwan Yeon

I enjoy researching Physical AI

About Me

I am a highly motivated and skilled professional with a passion for the intersection of machine learning and robotics. My experience encompasses both theoretical and experimental applications of learning algorithms in real-world autonomous driving services. I have contributed to advancing the field of motion prediction, planning, and control particularly in making Level 4 urban autonomous driving services publicly accessible in South Korea, while also making contributions to related academic fields. My ultimate research goal is to develop a scalable and interpretable robot learning algorithm.

I love traveling, hiking, and spending time at cozy cafés. I’m a naturally positive and warm-hearted person who enjoys finding humor and small joys even in busy days. I believe that having fun with what I do and staying curious makes both life and research more rewarding.

Projects

Learning-based Motion Planning

  • Designed high-level decision-making and trajectory selection logic, tailored for safety and efficiency in real-world autonomous navigation.
  • Developed a Transformer-based planning network for autonomous urban driving, capable of generating 1,024 optimized trajectory candidates within 30ms. Published at IROS.

Motion Prediction

  • Designed and implemented scene-, agent-, and goal-centric motion prediction models for autonomous vehicles to forecast surrounding agent trajectories in complex urban environments without relying on HD maps.
  • One of the developed models was presented at a CVPR Workshop.

Control and State Estimation

  • Engineered lateral and longitudinal control algorithms using Model Predictive Control (MPC), enabling smooth and responsive motion control in Level 4 autonomous vehicles.
  • Implemented vision-aided localization based on Extended Kalman Filter (EKF), improving robustness and positional accuracy under urban driving conditions.
  • Contributed to full-stack system integration across multiple vehicle platforms, including low-level control interfaces and deployment for public road operation.

Publications

  1. S. Moon, K. Yeon, H. Kim, S. -G. Jeong, J. Kim, “Who Should Have Been Focused: Transferring Attention- Based Knowledge from Future Observations for Trajectory Prediction”, International Conference on Pattern Recognition (ICPR), 2024 - Conference
  2. K. Yeon, H. Kim, S. -G. Jeong, “SpeedFormer: Learning Speed Profiles with Upper and Lower Boundary Constraints Based on Transformer,” IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2023 - Conference
  3. S. Moon, S. Lee, H. Woo, K. Yeon, H. Kim, S. -G Jeong, J. Kim, “RUFI: Reducing Uncertainty in behavior prediction with Future Information”, Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR Workshops), 2023 - Workshop
  4. J. Shin, K. Yeon, S. Kim, M. Sunwoo, M. Han, “Comparative Study of Markov Chain With Recurrent Neural Network for Short Term Velocity Prediction Implemented on an Embedded System,” IEEE Access, 2021 - Journal
  5. K. Min, K. Yeon, Y.Jo, “Vehicle Deceleration Prediction Based on Deep Neural Network at Braking Conditions”, International Journal of Automotive Technology (IJAT), 2020 - Journal
  6. K. Yeon, K. Min, J. Shin, M. Sunwoo, M. Han, “Ego-Vehicle Speed Prediction Using a Long Short-Term Memory Based Recurrent Neural Network”, International Journal of Automotive Technology (IJAT), 2019 - Journal, 99 times cited
  7. K. Min, K. Yeon, G. Sim, M. Sunwoo, “Prediction Algorithm for Decelerating Driving States Based on Driver Characteristics for Smart Regenerative Control of Electric Vehicles”, 8th Aachen Colloquium China Automobile and Engine Technology (IJAT), 2018 - Conference
  8. W. Kim, K. Yeon, S. -S. Lee, “Development of a Pitch Control Algorithm Through MRAC Based Longitu- dinal and Vertical Integrated Chassis Control”, Korean Society of Automotive Engineers (KSAE), 2018 - Poster

Patents

  1. K. Yeon, H. Kim, “Method, apparatus, and computer-readable medium for predicting a future trajectory of a target vehicle using movement information of one or more past surrounding vehicles”, Korean Patent, No. 1020220123725, 2025

  2. S. Jeong, K. Yeon, I. Bae, H. Kwon, “Method, apparatus, and recording medium for estimating the location and pose of a vehicle”, Korean Patent, No. 1020220022032, 2023