Qianfan Zhao

Papers

Total Citations

H-Index

About

Qianfan Zhao is a rising researcher in embodied AI and robotic visual navigation, whose work focuses on enabling robots to intelligently locate objects in real-world environments. His primary research areas include zero-shot object goal visual navigation, active visual learning, and semantic policy learning for robotics. Zhao’s most significant contribution is pioneering zero-shot object goal visual navigation (ZSON), which allows robots to locate “unseen” objects—those not encountered during training—by leveraging auxiliary semantic knowledge. His 2023 paper on this topic has garnered 36 citations, establishing him as a key contributor to this emerging field. He also developed the Semantic Policy Network, which further advances transferable navigation policies for novel objects (13 citations). Additionally, Zhao created a real 3D embodied dataset for robotic active visual learning (17 citations), addressing the critical gap between synthetic simulations and real-world deployment. This dataset enables robots to actively interact with environments to facilitate visual tasks, moving beyond static observations. Zhao’s work bridges the divide between controlled lab settings and practical household applications, making him a notable figure in the push toward truly autonomous home-assistance robots.