Yu-Ting Ko
Papers
2
Total Citations
48
H-Index
2
About
Yu-Ting Ko is a rising researcher in reinforcement learning and human-robot interaction, whose work bridges the gap between simulation and real-world robotic autonomy. Her key research areas include curriculum reinforcement learning, robot navigation, and federated manipulation learning. In her highly cited 2023 paper on curriculum reinforcement learning, Ko demonstrated how structured training curricula can dramatically accelerate learning convergence, enabling robots to transition from simple collision avoidance to navigating complex environments with movable obstacles—achieving 37 citations for advancing scalable training methodologies. Her work on Fed-HANet, with 11 citations, tackles the critical challenge of human-robot handovers by introducing a federated learning framework for visual grasping, allowing robots to learn robust, object-agnostic grasping policies across diverse environments without centralizing sensitive data. This innovation directly supports service robots in healthcare and logistics settings. Ko’s contributions are notable for their practical impact on real-world deployment, offering efficient, privacy-preserving solutions that enhance robot adaptability. Her research continues to shape how autonomous systems learn from interaction, making her a compelling voice in modern robotics and reinforcement learning.
Research Focus
Key Achievements
Top Papers
- 1
- 2Fed-HANet: Federated Visual Grasping Learning for Human Robot Handovers11 citations · 2023