Chengcheng Tang

Stanford University, Georgia Institute of Technology, META Health

Papers

Total Citations

H-Index

About

Chengcheng Tang is a researcher whose work sits at the compelling intersection of 3D geometry understanding, human-object interaction, and robotic manipulation. His most influential contribution, "Parsing Geometry Using Structure-Aware Shape Templates" (2018), garnered 37 citations and introduced a framework for decomposing man-made objects into structured, semantically meaningful parts — leveraging the inherent regularity of designed objects to improve geometric understanding. This work reflects a broader commitment to bridging visual perception and structural reasoning, with clear implications for computer vision and scene understanding. Tang's research extends naturally into the domain of hand-object interaction, most notably through his co-authorship of "ContactPose" (2020), a richly annotated dataset capturing grasp configurations alongside object contact patterns and hand pose — a resource that has become a valuable benchmark for the robotics and computer vision communities. More recently, his work on visual pressure estimation for soft robotic grippers (2022) demonstrates a growing interest in precision manipulation, proposing elegant vision-based solutions to the challenges posed by compliant, deformable robotic systems. Across these contributions, Tang demonstrates a distinctive ability to move fluidly between shape analysis, human motion understanding, and applied robotics — making him a researcher worth watching as these fields continue to converge.