Deyu Yin
Papers
1
Total Citations
21
H-Index
1
About
Deyu Yin is a researcher advancing the frontiers of autonomous navigation and 3D perception, with a primary focus on LiDAR odometry and unsupervised deep learning. His most cited work, "CAE-LO: LiDAR Odometry Leveraging Fully Unsupervised Convolutional Auto-Encoder for Interest Point Detection and Feature Description" (2020, 21 citations), introduces a novel framework that transforms raw 3D LiDAR data into compact 2D representations, enabling fully unsupervised learning for keypoint detection and feature matching. This contribution addresses a critical challenge in autonomous driving, robot navigation, and 3D mapping—achieving high-performance, easily adaptable odometry without the need for labeled data. By leveraging convolutional auto-encoders, Yin’s approach enhances robustness in complex environments, reducing reliance on handcrafted features. His work demonstrates how efficient data structuring and self-supervised techniques can simplify LiDAR-based systems while maintaining accuracy. With growing recognition in the field, Deyu Yin’s research continues to influence the development of scalable, unsupervised solutions for real-world spatial intelligence, making him a notable figure in modern robotics and autonomous systems.
Research Focus
Key Achievements
Top Papers
- 1