Zitian Wang

Beijing Technology and Business University, Beihang University

Papers

Total Citations

H-Index

About

Zitian Wang is a robotics and computer vision researcher whose work centers on multi-sensor fusion for state estimation and human-centric scene understanding. His most impactful contribution is the development of Lvio-Fusion, a self-adaptive, tightly coupled SLAM framework that fuses stereo camera, IMU, and LiDAR data using an actor-critic reinforcement learning method. This framework intelligently adjusts sensor weighting based on environmental conditions, addressing a critical challenge in mobile robotics—how to maintain robust localization when individual sensors degrade. With 48 citations, Lvio-Fusion has become a notable reference in adaptive sensor fusion. Wang has also advanced vision-language understanding through his work on human-centric relation segmentation, which tackles fine-grained tasks like recognizing "the book in the girl's left hand." This research pushes the boundaries of how robots interpret complex spatial and relational commands. By bridging robust state estimation with nuanced visual reasoning, Wang is contributing to the next generation of autonomous systems that can both navigate reliably and interact intelligently with human environments.