Jay Ming Wong

Papers

Total Citations

162

H-Index

About

Jay Ming Wong is a robotics researcher whose work lies at the intersection of deep learning, computer vision, and autonomous manipulation. His primary research areas include semantic segmentation, pose estimation, and dense scene reconstruction for robotic systems operating in unstructured environments. Wong’s most impactful contribution is the SegICP framework (151 citations), which integrates deep semantic segmentation with pose estimation to enable robots to perceive and manipulate task-relevant objects with greater speed and robustness in complex, real-world scenarios. He extended this work with SegICP-DSR, a real-time system that achieves millimeter-level pose accuracy (7.9 mm, σ=7.6 mm) and successfully identifies objects in cluttered settings. Wong has also explored the frontier of lifelong self-supervision in robotics, surveying how deep learning can be adapted for continuous, autonomous learning without human-labeled data. His research directly addresses critical bottlenecks in robotic manipulation competitions and industrial automation, demonstrating that perception systems can be both fast and reliable. With a focus on practical, deployable solutions, Wong’s work continues to shape how robots understand and interact with their environments.