Yuhao Shan
Papers
2
Total Citations
4
H-Index
2
About
Yuhao Shan is a researcher in computer and robot vision, with a focus on omnidirectional camera pose estimation and visual localization. His work addresses fundamental challenges in enabling robots to understand their position and orientation in space using spherical imaging. Shan’s key contributions include applying convolutional neural networks, such as PoseNet, to estimate the pose of omnidirectional cameras from equirectangular and perspective mosaic images. He further advanced the field by developing a CNN-LSTM network that leverages multiple spherical images from neighboring places to resolve visual ambiguity—a common failure point in robot localization when scenes appear similar. Although his most-cited papers currently hold 2 citations each, they represent early, innovative steps toward robust, vision-based navigation for autonomous systems. By tackling the problem of confusion in visually repetitive environments, Shan’s work lays important groundwork for more reliable robot localization in real-world settings, making his research of particular interest to students and engineers working at the intersection of deep learning and spatial perception.
Research Focus
Key Achievements
Top Papers
- 1
- 2