Max Robotics — Global Service Robot Directory + US Certification

About

Yuchao Dai is a computer vision researcher whose work spans 3D scene understanding, deep learning, and camera systems, with particular expertise in point cloud processing, stereo matching, and camera modeling. His most influential contribution, a 2020 survey on deep learning-based point cloud registration, has garnered 149 citations and has become an essential reference for researchers working on 3D reconstruction, SLAM, and autonomous driving — areas where aligning spatial data accurately is critical. Dai has also advanced stereo matching through his Cross-Form Pyramid Network (CFP-Net), a novel architecture designed to improve disparity estimation for real-world applications including robotics and self-driving vehicles. A distinctive thread in his research involves rolling shutter cameras — a pervasive yet technically challenging imaging modality — with recent work addressing their modeling, optimization, and learning-based correction, including solutions for rolling shutter stereo rigs using differential Structure from Motion. Earlier in his career, Dai tackled the practical problem of RGB-D camera calibration, proposing scene-constraint-based extrinsic methods relevant to augmented reality and robotics. Together, his publications reflect a researcher dedicated to bridging geometric fundamentals with modern deep learning to solve demanding real-world vision challenges.