Ziming Fan

Papers

Total Citations

H-Index

About

Ziming Fan is a rising researcher in computer vision and robotics, whose work centers on category-level 6D object pose estimation—a critical capability for enabling robots to interact with unseen objects in unstructured environments. His most cited paper, “Category-Level 6D Pose Estimation Using Geometry-Guided Instance-Aware Prior and Multi-Stage Reconstruction” (2023, 8 citations), tackles the fundamental challenge of predicting the translation and rotation of arbitrary object instances from known categories without requiring per-instance 3D models. Fan’s key contribution lies in developing a geometry-guided, instance-aware prior that leverages shape reconstruction across multiple stages, significantly improving pose accuracy for diverse, unseen objects. This approach directly advances applications in robotic manipulation, augmented reality, and 3D scene understanding, where robust, generalizable pose estimation is essential. Though early in his career, Fan’s work has already garnered attention for its novel integration of geometric reasoning with learning-based reconstruction, marking him as a promising contributor to the field. His research bridges the gap between category-level generalization and precise spatial reasoning, offering a pathway toward more adaptable and intelligent robotic systems.