Hao-Shu Fang
Scholar

Hao-Shu Fang

Google Scholar ID: OT16tRAAAAAJ
Massachusetts Institute of Technology
Robotic ManipulationRobot LearningComputer Vision
Homepage↗Google Scholar↗
Citations & Impact
All-time
Citations
9,515
 
H-index
29
 
i10-index
38
 
Publications
20
 
Co-authors
0
 
Contact
Emailfhaoshu@gmail.comCVOpen ↗GitHubOpen ↗
Publications
7 items
Asynchronous Multimodal Diffusion Policy Composition via Latency-Aware Guidance Fusion
2026
Cited
0
Multi-Camera View Scaling for Data-Efficient Robot Imitation Learning
2026
Cited
0
ExoGS: A 4D Real-to-Sim-to-Real Framework for Scalable Manipulation Data Collection
2026
Cited
0
DEXOP: A Device for Robotic Transfer of Dexterous Human Manipulation
2025
Cited
0
VQ-VLA: Improving Vision-Language-Action Models via Scaling Vector-Quantized Action Tokenizers
2025
Cited
0
Dense Policy: Bidirectional Autoregressive Learning of Actions
2025
Cited
0
AirExo-2: Scaling up Generalizable Robotic Imitation Learning with Low-Cost Exoskeletons
2025
Cited
0
Resume
Academic Achievements
  • Multiple papers accepted and awarded at top conferences, such as DEXOP winning the Best Paper Award at RSS 2025 dexterous manipulation workshop; EyeSight Hand paper being a Best RoboCup Paper Finalist at IROS 2024; RT-X paper winning the Best Paper Award at ICRA 2024; AnyGrasp winning the Best Paper Award at IROS 2023 workshop. Also, received ByteDance Fellowship, Microsoft Research Asia Fellowship, and Baidu Fellowship.
Research Experience
  • Working at MIT CSAIL with Prof. Pulkit Agrawal and Prof. Edward Adelson.
Education
  • Ph.D. from Shanghai Jiao Tong University, supervised by Prof. Cewu Lu.
Background
  • Postdoc researcher at MIT CSAIL, focusing on General Robotic Manipulation. Will join the University of Maryland, College Park (UMD) as an Assistant Professor of Computer Science in 2026.
Miscellany
  • Welcomes students with ME and EE backgrounds to join his team.