Papers
60
Total Citations
4,009
H-Index
22
About
Sehoon Ha is a prominent researcher at the intersection of robotics, deep reinforcement learning, and computational design, whose work has fundamentally advanced how robots learn to move and adapt in the real world. Best known for his contributions to the landmark "Soft Actor-Critic Algorithms and Applications" paper (2018, nearly 2,000 citations), Ha helped establish SAC as one of the most widely adopted model-free deep RL algorithms, addressing critical challenges of sample efficiency and hyperparameter sensitivity. His research on legged locomotion has been equally impactful, with multiple influential works demonstrating how robots can learn robust walking behaviors through deep RL, both in simulation and directly in physical environments with minimal human intervention. Ha has also pioneered computational co-design methods that jointly optimize robot morphology and motion trajectories, enabling automated design of robotic systems from high-level specifications. More recently, his work on meta-learning for fast locomotion adaptation and zero-shot semantic navigation using vision-language models highlights his range across core robotics challenges. His contributions to the DART simulation toolkit further reflect his commitment to open, community-driven research infrastructure, making him a highly influential figure in modern robotics and AI.
Research Focus
Key Achievements
Top Papers
- 1Soft Actor-Critic Algorithms and Applications1,952 citations · 2018
- 2Learning to Walk Via Deep Reinforcement Learning434 citations · 2019
- 3DART: Dynamic Animation and Robotics Toolkit272 citations · 2018
- 4
- 5VLFM: Vision-Language Frontier Maps for Zero-Shot Semantic Navigation93 citations · 2024
- 6
- 7Learning Fast Adaptation With Meta Strategy Optimization84 citations · 2020
- 8Learning to Walk in the Real World with Minimal Human Effort76 citations · 2020
- 9
- 10