Laiyan Ding

Chinese University of Hong Kong, Shenzhen

Papers

Total Citations

H-Index

About

Laiyan Ding is an emerging researcher at the forefront of computer vision and robotics, with a specialized focus on depth estimation and stereo matching. Their most notable contribution, **DEFOM-Stereo: Depth Foundation Model Based Stereo Matching** (2025), represents a significant advancement in the field by addressing longstanding challenges in metric depth estimation, including occlusion handling and accurate disparity estimation in non-textured regions. By innovatively integrating monocular relative depth estimation's remarkable generalization capabilities into binocular stereo matching frameworks, Ding's work bridges a critical gap between these two complementary approaches to 3D scene understanding. The paper has already garnered 19 citations within its first year of publication — a strong indicator of immediate impact within the computer vision community. This work holds particular relevance for robotics applications where precise, real-world depth perception is essential for autonomous navigation and interaction. Ding's research reflects a broader trend of leveraging large-scale foundation models to enhance traditional geometric computer vision pipelines, positioning their work at a compelling intersection of deep learning and classical 3D vision. Their early-career output suggests a promising trajectory in the field.