Tianyu Zhang

State Key Laboratory of Transducer Technology

Papers

Total Citations

H-Index

About

Tianyu Zhang is an emerging researcher in the field of robotics and autonomous systems, with a focused expertise in visual-inertial odometry (VIO) and intelligent robot localization. His most notable work, "Self-supervised Scale Recovery for Decoupled Visual-inertial Odometry" (2023), addresses one of the fundamental challenges in self-supervised VIO systems: the effective integration of inertial measurement data to enable accurate absolute scale recovery. Rather than treating inertial information as generic input — a limitation common to prior approaches — Zhang's framework decouples the visual and inertial components to preserve and leverage the inherent physical properties of inertial sensors, enabling more robust and metrically accurate pose estimation. This contribution represents a meaningful step forward in making self-supervised localization systems practical for real-world robotic deployment, where ground-truth supervision is often unavailable. While his work is in its early stages of accumulating citations, with 3 citations to date, the problem it addresses — reliable, scalable localization for autonomous robots — sits at the heart of modern robotics research. Zhang's work positions him as a promising contributor to the growing community advancing self-supervised perception for intelligent systems.