Ziyu Guan

Papers

Total Citations

H-Index

About

Ziyu Guan is a researcher advancing the frontiers of computer vision and medical image analysis, with a focus on self-supervised learning and 3D understanding. His work addresses critical challenges in data scarcity and annotation costs, particularly for high-stakes applications like surgical robotics and autonomous systems. In his highly cited 2023 paper, "A General Global and Local Pre-Training Framework for 3D Medical Image Segmentation," Guan tackles the persistent bottleneck of limited medical imaging data. By introducing a novel self-supervised framework that learns both global and local features from unlabeled CT scans, his method enables robust segmentation models without extensive manual annotations—a breakthrough for clinical deployment. This work has garnered 5 citations, reflecting its immediate impact on the medical imaging community. More recently, in his 2025 study "Learning Cross-View Consistent 3D Keypoints for Object 6D Pose Estimation," Guan addresses the reliance on expensive labeled data for 6D pose estimation. By learning consistent 3D keypoints across multiple views, his approach reduces the need for synthetic or manually annotated datasets, offering a scalable solution for augmented reality and robotic manipulation. With 3 citations already, this work signals a growing influence. Guan’s contributions are notable for their practical focus on reducing data dependencies, making advanced vision systems more accessible and deployable in real-world environments.