Papers
6
Total Citations
86
H-Index
4
About
En Yen Puang is a robotics researcher whose work sits at the intersection of computer vision, deep learning, and autonomous manipulation. His research focuses on enabling robots to perceive and interact with their environments more intelligently, with particular emphasis on visual servoing, keypoint-based representations, and reinforcement learning for manipulation tasks. His most influential contribution, KOVIS (2020, 52 citations), introduced a calibration-free visual servoing framework that leverages keypoint detection and achieves zero-shot sim-to-real transfer — allowing robots trained entirely in simulation to operate effectively in the real world without additional calibration. This work addressed a critical bottleneck in deploying learning-based robotic systems practically. Beyond visual servoing, Puang has contributed to motion planning through visual repetition sampling, 3D point cloud representation learning with lightweight self-attention architectures, and end-to-end reinforcement learning frameworks using robust keypoint state representations. His work with Team NimbRo at MBZIRC 2017, where the team won Challenge 2 for autonomous valve-stem turning, demonstrates his strength in real-world competitive robotics. His more recent research explores tactile-informed grasp stability using transformer-based control policies, reflecting a growing interest in multimodal robot dexterity. Collectively, his work advances the reliability and generalizability of autonomous robotic manipulation.
Research Focus
Key Achievements
Top Papers
- 1
- 2Team NimbRo at MBZIRC 2017: Autonomous valve stem turning using a wrench16 citations · 2018
- 3Visual Repetition Sampling for Robot Manipulation Planning6 citations · 2019
- 4
- 5
- 6