Hyung-Suk Yoon
Papers
2
Total Citations
16
H-Index
2
About
Hyung-Suk Yoon is a leading researcher in autonomous robot navigation, specializing in traversability estimation and robust decision-making for unstructured environments. His work addresses a critical challenge in field robotics: enabling robots to safely navigate off-road terrains where traditional mapping fails. Yoon’s most impactful contribution is his 2024 paper on adaptive robot traversability estimation, which introduces a self-supervised online continual learning framework. This approach allows robots to continuously update their understanding of terrain traversability from real-time experience, achieving 13 citations in its first year—a strong indicator of its significance to the community. In his 2022 work, UNICON, Yoon tackled the overconfidence problem in deep reinforcement learning agents by developing an uncertainty-conditioned policy that adapts behavior in unfamiliar scenarios. This innovation is vital for safety-critical applications, ensuring robots act cautiously when facing novel states. Yoon’s research bridges the gap between theoretical reinforcement learning and practical deployment, earning recognition for advancing autonomous navigation in challenging, real-world conditions.
Research Focus
Key Achievements
Top Papers
- 1
- 2