Mike Pan

Papers

Total Citations

278

H-Index

About

Mike Pan is a researcher whose work sits at the intersection of computer vision and robotic manipulation, with a particular focus on solving some of the most challenging perception problems in real-world environments. He is best known for his landmark contributions to the problem of transparent object understanding — a notoriously difficult challenge in robotics, as standard 3D depth sensors routinely fail to capture accurate geometry for glass, plastic, and other see-through materials. Pan's most recognized work, *ClearGrasp*, introduced a groundbreaking approach to estimating the 3D shape of transparent objects to enable reliable robotic grasping. By leveraging deep learning to infer surface normals, occlusion boundaries, and contact edges, the system reconstructs accurate depth information where conventional sensors fall short. The 2020 publication of this work has amassed 258 citations, reflecting its significant influence on the robotics and computer vision communities and cementing it as a key reference for researchers tackling transparent object manipulation. An earlier version of the work presented in 2019 further demonstrates Pan's sustained commitment to advancing this field. His research has meaningful implications for warehouse automation, assistive robotics, and any domain requiring reliable interaction with everyday transparent objects.