Papers
81
Total Citations
4,277
H-Index
31
About
Marc Pollefeys is a distinguished computer vision and robotics researcher whose work spans visual localization, autonomous navigation, dense 3D reconstruction, and multi-sensor fusion. He has made foundational contributions to enabling machines to understand and navigate their environments, with particular impact in structure-from-motion, SLAM, and camera pose estimation. His work on PX4, the open-source robotics middleware for embedded platforms, has become a cornerstone of the autonomous systems community, accumulating nearly 700 citations and powering countless real-world drone applications. His investigations into the limitations of CNN-based camera pose regression (401 citations) have shaped how researchers critically evaluate deep learning approaches to visual localization. Pollefeys has also advanced micro aerial vehicle autonomy, developing vision-controlled flying robots capable of GPS-denied navigation and mapping (297 citations), and contributed key tools including an open-source optical flow camera and the CamOdoCal multi-camera calibration system. His more recent research on LiDAR-inertial-camera fusion and the HoloLens 2 Research Mode platform reflects a sustained commitment to bridging computer vision with mixed reality and robotics. Collectively, his publications have garnered thousands of citations, cementing his reputation as a seminal figure in applied computer vision.
Research Focus
Key Achievements
Top Papers
- 1
- 2Understanding the Limitations of CNN-Based Absolute Camera Pose Regression401 citations · 2019
- 3
- 4
- 5
- 6
- 7Semantic Match Consistency for Long-Term Visual Localization150 citations · 2018
- 8Robust Dense Mapping for Large-Scale Dynamic Environments147 citations · 2018
- 9
- 10HoloLens 2 Research Mode as a Tool for Computer Vision Research118 citations · 2020