Papers
1
Total Citations
2
H-Index
1
About
Li Dong is a rising computer vision researcher whose work focuses on real-time 3D scene understanding, particularly through plane segmentation and geometric deep learning. His most notable contribution is the development of FastPlane (2024), a fully convolutional network that achieves real-time 3D plane segmentation—a challenging task involving the simultaneous detection of planar regions and prediction of their spatial orientation. By reframing this as a multi-task learning problem with separate network branches, Dong’s approach enables efficient, end-to-end inference without sacrificing accuracy. Though early in his career, his work addresses a critical bottleneck in robotics, augmented reality, and autonomous navigation: the need for fast, reliable geometric parsing of cluttered environments. FastPlane’s architecture has already garnered attention for its balance of speed and precision, laying groundwork for future systems that must interpret 3D scenes in real time. As his citation count grows, Dong is establishing himself as a key voice in bridging deep learning with practical 3D perception.
Research Focus
Key Achievements
Top Papers
- 1