Changxi Zheng

Papers

Total Citations

203

H-Index

About

Changxi Zheng is a researcher whose work sits at the intersection of robotics, machine learning, and autonomous manipulation, with a particular focus on enabling robots to interact intelligently with their physical environments. His most recognized contributions center on the challenge of **object placement** — teaching robots not merely to set objects down stably, but to reason about semantically appropriate locations and orientations within real-world scenes. This deceptively complex problem requires integrating physical stability constraints with learned contextual preferences, such as understanding that a plate belongs vertically in a dish rack rather than laid flat on a counter. Zheng's foundational 2012 paper, "Learning to Place New Objects in a Scene," has accumulated 147 citations, establishing it as a key reference in the robot manipulation and scene understanding literature. His body of work on this topic spans multiple publication venues and iterations, reflecting a sustained and rigorous development of the core ideas. By framing object placement as a learning problem — one where robots generalize from observed examples to novel objects and environments — Zheng helped advance the broader goal of capable, adaptable personal robotics. His research remains relevant to ongoing efforts in household automation, assistive robotics, and embodied AI.