Papers
7
Total Citations
452
H-Index
6
About
Fei-Peng Tian is a computer vision and robotics researcher whose work centers on depth perception, autonomous navigation, and camera relocalization. He is perhaps best known for his highly influential research on depth completion, particularly his "Learning Guided Convolutional Network for Depth Completion," which has accumulated nearly 300 citations and addresses the critical challenge of transforming sparse LiDAR measurements into dense depth maps using RGB image guidance — a problem with direct implications for autonomous driving and robotic systems. His 2019 work on learned map prediction for mobile robot exploration, cited over 100 times, demonstrates his ability to apply deep and reinforcement learning to classical robotics problems, improving how autonomous robots navigate and reconstruct unknown environments. Tian has also made notable contributions to active camera relocalization, developing methods that eliminate the need for cumbersome hand-eye calibration and extend to RGBD and line-segment-based approaches. His additional work on computational rephotography brings practical computer vision solutions to mobile platforms. Collectively, his research portfolio reflects a consistent focus on bridging the gap between theoretical machine learning advances and real-world robotic and imaging applications, making him a meaningful contributor to the autonomous systems research community.
Research Focus
Key Achievements
Top Papers
- 1Learning Guided Convolutional Network for Depth Completion273 citations · 2020
- 2Learned Map Prediction for Enhanced Mobile Robot Exploration101 citations · 2019
- 3
- 4Learning Guided Convolutional Network for Depth Completion26 citations · 2019
- 5Fast and Reliable Computational Rephotography on Mobile Device15 citations · 2018
- 6Active Camera Relocalization with RGBD Camera from a Single 2D Image8 citations · 2018
- 7