Max Robotics — Global Service Robot Directory + US Certification

About

Chilam Cheang is a researcher working at the intersection of computer vision, robotics, and multimodal AI, with particular expertise in 6D object pose estimation, robotic grasping, and vision-language models for robot manipulation. His most influential contribution, "SAR-Net: Shape Alignment and Recovery Network for Category-level 6D Object Pose and Size Estimation" (2022), has garnered 86 citations and addresses the challenging problem of estimating object pose and size from point cloud data without requiring annotated real-world training data — a significant practical advancement for robotics applications. His earlier work, "DONet" (2021), laid groundwork in depth-based pose estimation using purely geometric information. Cheang has also pushed boundaries in human-robot interaction, exploring how freehand sketches and natural language instructions can guide robotic grasping systems. More recently, his work on vision-language foundation models — including the generalist robot agent GR-2, pre-trained on 38 million video clips — reflects a growing focus on scalable, generalizable robot learning. Collectively, his research demonstrates a clear trajectory toward enabling robots to understand and interact with the physical world through rich multimodal perception, making meaningful contributions to the field of embodied AI.