About

Tianmin Shu is a researcher at the forefront of socially intelligent robotics and human-robot interaction (HRI), with a body of work that bridges cognitive science, computer vision, and autonomous systems. His research centers on enabling robots to understand, model, and participate in complex social dynamics — a challenge he approaches through probabilistic reasoning, structured grammar models, and theory of mind. Shu's most impactful contributions include pioneering the use of spatiotemporal AND-OR graphs (ST-AOGs) to learn social affordance grammars directly from RGB-D videos, allowing robots to infer and replicate human interaction patterns in real time (garnering nearly 40 citations). His work on joint mind modeling advances robots' ability to generate meaningful explanations during collaborative tasks by reasoning about partners' goals, beliefs, and desires. More recently, he has formalized rich sociological theories of human interaction within Markov Decision Process (MDP) frameworks, pushing robots toward genuine social reasoning rather than narrow task execution. With contributions spanning capability calibration for collaborative robots and neurally-guided assistance for home robots, Shu's research consistently addresses a critical gap: giving machines the social fluency needed to be true human partners. His work has accumulated over 130 citations across a growing publication record.

Research Focus

Key Achievements

6
H-Index
10
Papers
130
Total Citations
13
Avg Citations/Paper
🏆 Most Cited Paper
Learning social affordance grammar from videos: Transferring human interactions to human-robot interactions
39 citations · 2017
📈 Most Prolific Year: 2017 (2 Papers)
🤝 Key Collaborators: 18
🏛 Institutions: University of California, Los Angeles, Massachusetts Institute of Technology, MIT Art, Design and Technology University

Top Papers

  1. 1
  2. 2
  3. 3
  4. 4
  5. 5
  6. 6
  7. 7
  8. 8
  9. 9
  10. 10

Key Collaborators

Contact & Links

Available for collaboration
Content generated · 14 days ago