Haozhe Xie

Harbin Institute of Technology, Group Sense (China)

Papers

Total Citations

425

H-Index

About

Haozhe Xie is a leading researcher in 3D computer vision, with a primary focus on point cloud completion and 3D object reconstruction. His most influential contribution is the GRNet (Gridding Residual Network), introduced in 2020, which addresses the critical challenge of estimating complete 3D point clouds from partial or incomplete data. Unlike mainstream methods that rely on Multi-layer Perceptrons (MLPs) and often lose structural details, GRNet employs a novel gridding operation to preserve fine-grained geometric information, achieving dense and high-fidelity point cloud completion. This work has garnered over 375 citations, underscoring its impact on the field. Xie has also advanced 3D object reconstruction from stereo images, developing techniques that move beyond simple template matching to improve generalization and reconstruction quality. His research is widely cited in robotics, autonomous navigation, and augmented reality applications. By tackling fundamental limitations in 3D data processing, Haozhe Xie continues to shape the future of how machines perceive and reconstruct the three-dimensional world.