About Me
Qianxu Wang
I am a second-year PhD student in Computer Science at Cornell University, advised by Prof. Kuan Fang. I am interested in robot manipulation that generalizes, including cross-embodiment transfer, semantic 3D representations for one-shot dexterous manipulation, and self-supervised adaptation.
Previously, I graduated with highest honors from Peking University with a bachelor’s degree in Computer Science. I was fortunate to work with Prof. Yixin Zhu at Peking University, Prof. Jeannette Bohg at Stanford, and Prof. Leonidas J. Guibas at Stanford.
Research
Generalization in robot manipulation.
My research focuses on generalization in robot manipulation across objects, environments, and embodiments. I develop visual and interaction representations that help robots transfer skills beyond the settings in which they were learned, drawing on pretrained vision models and prior manipulation experience. I am also interested in self-supervised adaptation at deployment, where robots use observations of new environments to refine their perception and behavior.
Feel free (and please do!) to reach out if you have any questions, comments about my research, or anything you'd like to discuss or share with me!
Publications
Selected publications.
Conference on Robot Learning (CoRL) 2024
Neural Attention Field: Emerging Point Relevance in 3D Scenes for One-Shot Dexterous Grasping
More publications can be found on Google Scholar.
Misc
A few fun facts.
I enjoy doing nails (and can do yours, too). I love photography and play squash.
Photography
