Jason Xinyu Liu


Hello and Welcome!

I am a computer science Ph.D. student at Brown University. I am fortunately to be advised by Prof. Stefanie Tellex and work with Prof. George Konidaris. I received my bachelor of science degree from Berkeley EECS . I am grateful for the supports of the NSF Graduate Research Fellowship Program and Jack Kent Cooke Foundation Graduate Scholarship.

Research Interests

I am interested in building intelligent robots that can complete a broad range of long-horizon tasks in diverse environments provided with sparse reward signals. To that end, I am developing intelligent systems that can leverage structure in task specifications and environments for efficient planning and learning. The approaches I am currently investigating are to apply RL algorithms to solve robotic tasks specified by natural language with specific structure.


Skill Transfer for Temporally-Extended Task Specifications. [paper]
Jason Xinyu Liu∗, Ankit Shah∗, Eric Rosen, George Konidaris, Stefanie Tellex.
preprint 2022.

Leveraging Temporal Structure in Task Specifications for POMDP Planning.
Xinyu Liu, Eric Rosen, Ankit Shah, Suchen Zheng, Tyler Edwards, George Konidaris, Stefanie Tellex.
In Submission.

Generalizing to New Domains by Mapping Natural Language to Lifted LTL. [paper]
Eric Hsiung, Hiloni Mehta, Junchi Chu, Xinyu Liu, Roma Patel, Stefanie Tellex, George Konidaris.
IEEE International Conference on Robotics and Automation (ICRA) 2022.

Leveraging Temporal Structure in Safety-Critical Task Specifications for POMDP Planning. [paper, poster]
Xinyu Liu, Eric Rosen, Suchen Zheng, George Konidaris, Stefanie Tellex.
Robotics: Science and Systems (RSS) 2021 Workshop Robotics for People.

Dialogue Object Search. [paper]
Monica Roy, Kaiyu Zheng, Xinyu Liu, Stefanie Tellex.
Robotics: Science and Systems (RSS) 2021 Workshop Robotics for People.

Specificity-Controlled Video Captioning. [paper]
Xinyu Liu, Ellie Pavlick, Daniel Ritchie, George Konidaris, Stefanie Tellex.
Technical Report 2019.

Dex-Net 3.0: Computing Robust Vacuum Suction Grasp Targets in Point Clouds Using a New Analytic Model and Deep Learning. [paper, site]
Jeffrey Mahler, Matthew Matl, Xinyu Liu, Albert Li, David Gealy, Ken Goldberg.
IEEE International Conference on Robotics and Automation (ICRA) 2018.

Detecting Phone Theft Using Machine Learning. [paper]
Xinyu Liu, David Wagner, Serge Egelman.
International Conference on Information Science and System (ICISS) 2018.

Dex-Net 2.0: Deep Learning to Plan Robust Grasps with Synthetic Point Clouds and Analytic Grasp Metrics. [paper, site]
Jeffrey Mahler, Jacky Liang, Sherdil Niyaz, Michael Laskey, Richard Doan, Xinyu Liu, Juan Aparicio Ojea, Ken Goldberg.
Robotics: Science and Systems (RSS) 2017.

Fun Outside Work

I love reading books, playing basketball, hiking, camping and movies directed by Christopher Nolan. When I read, I listen to guitar or Jazz BGM. The books I recently read are The Road Less Traveled, The Autobiography of Benjamin Franklin, and How We Think by John Dewey.


Lmk if you'd like to chat. Always up to have fun conversations and make new friends!