About

Wei Feng is a researcher whose work sits at the intersection of computer vision, robotics, and deep learning, with particular expertise in depth perception, autonomous navigation, and camera localization. His most influential contribution, "Learning Guided Convolutional Network for Depth Completion" (2020, 273 citations), tackled a fundamental challenge in autonomous driving by developing a deep learning framework that transforms sparse LiDAR measurements into dense depth maps using synchronized RGB imagery — a breakthrough that has become widely referenced across the robotics and self-driving communities. His 2019 work on map prediction for mobile robot exploration (101 citations) further demonstrated his ability to leverage learned approaches to surpass traditional geometric methods in autonomous indoor navigation. Feng's research portfolio also reflects a sustained interest in active camera relocalization, hand-eye calibration, and human-robot interaction, including early contributions in RGB-D-based hand posture recognition. His 2018 paper on active relocalization from a single reference image introduced a novel, calibration-free paradigm that challenged conventional assumptions in the field. More recently, his foray into material intelligence signals an expanding interdisciplinary vision. Collectively, Feng's body of work — spanning over 400 citations — reflects a researcher continuously pushing the boundaries of intelligent perception and autonomous systems.

Research Focus

Key Achievements

6
H-Index
8
Papers
436
Total Citations
55
Avg Citations/Paper
🏆 Most Cited Paper
Learning Guided Convolutional Network for Depth Completion
273 citations · 2020
📈 Most Prolific Year: 2018 (3 Papers)
🤝 Key Collaborators: 40
🏛 Institutions: State Administration of Cultural Heritage, Tianjin University, Chinese Academy of Cultural Heritage, Chinese Academy of Sciences, Shenzhen Institute of Information Technology

Top Papers

  1. 1
  2. 2
  3. 3
  4. 4
  5. 5
  6. 6
  7. 7
  8. 8

Key Collaborators

Contact & Links

Available for collaboration
Content generated · 14 days ago