Papers
2
Total Citations
5
H-Index
1
About
Ran Yi is a researcher at the forefront of robotics and 3D computer vision, whose work bridges the gap between classical motion planning and modern machine learning. Her key research areas include sampling-based motion planning for high-degree-of-freedom robots and point cloud completion for 3D scene understanding. Yi’s major contribution in robotics is a novel configuration-space decomposition scheme that enables learning-based collision checking, significantly accelerating motion planning for complex robotic systems. This work, published in 2019, has garnered 4 citations and represents an important step toward integrating learned classifiers into traditional sampling-based methods. More recently, Yi has advanced 3D vision with DT-Net, a point cloud completion framework that introduces a neighboring adaptive denoiser and a splitting-based upsampling transformer. This 2025 paper, already accumulating 1 citation, addresses the challenging problem of inferring missing 3D object geometry from partial observations—a critical capability for autonomous navigation and manipulation. By combining denoising with transformer-based upsampling, DT-Net achieves state-of-the-art performance in reconstructing complete, high-fidelity point clouds. Yi’s work exemplifies how deep learning can enhance both robotic planning and 3D perception, making her a rising voice in these interconnected fields.
Research Focus
Key Achievements
Top Papers
- 1
- 2