Yu‐Kun Lai
Papers
13
Total Citations
264
H-Index
9
About
Yu-Kun Lai is a prominent researcher whose work spans robotics, computer vision, and machine learning, with particular expertise in robotic manipulation, 3D scene reconstruction, and autonomous navigation. His research bridges the gap between low-level motion control and high-level symbolic reasoning, most notably through his influential work on hierarchical reinforcement learning for multistep robotic manipulation tasks such as block stacking and parts assembly, which has garnered 88 citations and represents a significant advance in autonomous robotics. Lai has made substantial contributions to 3D environment understanding, including multi-sensor dense scene reconstruction through HeteroFusion and noise-resilient panoramic reconstruction using RGB-D cameras. His work on point cloud-based place recognition (TransLoc3D) addresses critical challenges in autonomous driving and robot navigation. Lai also actively shapes the research community through survey papers on Object Goal Navigation and deep robotic affordance learning, providing valuable syntheses for emerging fields. His open-source multi-goal reinforcement learning environment further demonstrates a commitment to accessible, reproducible research. Collectively accumulating over 250 citations, his body of work positions him as a versatile and impactful contributor to embodied AI and intelligent robotic systems.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3
- 4A Survey of Object Goal Navigation21 citations · 2024
- 5
- 6HeteroFusion: Dense Scene Reconstruction Integrating Multi-Sensors20 citations · 2019
- 7
- 8
- 9
- 10