Youqi Liao

Wuhan University, Wuhan University of Technology

Papers

Total Citations

H-Index

About

Youqi Liao’s research sits at the intersection of robotics, autonomous driving, and multi-modal perception, with a sharp focus on making machines see and fuse data with unprecedented precision. His core contributions lie in sensor calibration, semantic understanding, and cross-modal registration — all critical for real-world robotic systems operating in dynamic urban environments. In his highly cited work “SE-Calib” (42 citations), Liao introduced a semantic edge-based method for online LiDAR-camera boresight calibration, eliminating the need for artificial targets and enabling robust, continuous alignment in unstructured scenes. This breakthrough directly addresses a fundamental bottleneck in sensor fusion for earth observation and autonomous navigation. Building on this, his “Mobile-Seed” framework (22 citations) achieves joint semantic segmentation and boundary detection on edge computing units, delivering the sharp, real-time semantic maps essential for grasping, manipulation, and online calibration. Most recently, “CoFiI2P” (21 citations) tackles the notoriously difficult problem of image-to-point cloud registration by introducing a coarse-to-fine correspondence pipeline that prioritizes global alignment before local refinement. Together, Liao’s work forms a cohesive toolkit for enabling robots to perceive, align, and act with reliability — earning him recognition as a rising leader in embodied perception and sensor fusion.