About

Lap-Fai Yu is a leading researcher at the intersection of computer vision, robotics, and human-robot interaction. His work spans scene understanding, assistive robotics, and affordance reasoning, with a particular focus on creating intelligent systems that can perceive, navigate, and interact with complex environments. Yu’s most impactful contribution is the SceneNN dataset (346 citations), a richly annotated RGB-D scene dataset that has become a foundational resource for 3D scene understanding and robotics research. He has also pioneered assistive technology for the blind and visually impaired, developing a deep learning-based robotic guide dog that learns trail-following behaviors from both virtual and real-world environments (73 citations). His innovative work on liquid containability reasoning introduced a physics-based approach to affordance prediction, enabling robots to understand object functionality beyond simple geometry. More recently, Yu has advanced human-robot collaboration through federated learning for grasping in handover tasks and designed frameworks for human-robot teaming in navigation among movable obstacles. His research on human-robot coexistence spaces and mixed-reality worker-drone interaction demonstrates a commitment to safe, effective human-robot collaboration in real-world settings.

Research Focus

Key Achievements

6
H-Index
8
Papers
479
Total Citations
60
Avg Citations/Paper
🏆 Most Cited Paper
SceneNN: A Scene Meshes Dataset with aNNotations
346 citations · 2016
📈 Most Prolific Year: 2016 (1 Papers)
🤝 Key Collaborators: 38
🏛 Institutions: Singapore University of Technology and Design, University of Massachusetts Boston, George Mason University

Top Papers

  1. 1
  2. 2
  3. 3
  4. 4
  5. 5
  6. 6
  7. 7
  8. 8

Key Collaborators

Contact & Links

Available for collaboration
Content generated · 13 days ago