Ji‐Xiang Du

About

Dr. Ji-Xiang Du is a leading researcher at the intersection of computer vision, robotics, and 3D perception, with a primary focus on point cloud analysis and 6D object pose estimation. His most impactful contribution is the comprehensive survey "Graph Neural Networks in Point Clouds: A Survey" (2024, 31 citations), which systematically maps the integration of GNNs with irregular 3D data—a cornerstone for autonomous driving and augmented reality. Dr. Du has advanced 3D object detection through novel attention mechanisms, as demonstrated in his work on "Extracting geometric and semantic point cloud features with gateway attention" (2023, 16 citations), which enhances spatial reasoning. In robotics, he developed an improved RRT* algorithm for robotic arm path planning (2022, 13 citations), addressing efficiency and path quality. His recent series on monocular 6D pose estimation—including spatial-temporal consistency learning (2024) and dual-branch PnP networks (2023)—pushes boundaries in single-image 3D localization, crucial for AR and manipulation. With a growing citation footprint, Dr. Du’s work bridges geometric deep learning and practical robotics, establishing him as a key innovator in 3D scene understanding.

Research Focus

Key Achievements

Top Papers

1. 1
   Graph Neural Networks in Point Clouds: A Survey

   31 citations · 2024
2. 2
   Extracting geometric and semantic point cloud features with gateway attention for accurate 3D object detection

   16 citations · 2023
3. 3
   Path Planning Method of Robot Arm Based on Improved RRT* Algorithm

   13 citations · 2022
4. 4
   Spatial and temporal consistency learning for monocular 6D pose estimation

   4 citations · 2024
5. 5
   Dual Branch PnP Based Network for Monocular 6D Pose Estimation

   2 citations · 2023
6. 6
   Images, normal maps and point clouds fusion decoder for 6D pose estimation

   2 citations · 2025

Key Collaborators

Contact & Links