Andrew Fitzgibbon

Papers

Total Citations

H-Index

About

Andrew Fitzgibbon is a leading figure in computer vision and machine learning, renowned for his foundational work in 3D reconstruction, camera calibration, and human pose estimation. His early research, including the development of the "Rθ mapping" as an extension of the Hough transform, laid groundwork for robust geometric feature detection. Fitzgibbon made seminal contributions to structure from motion and non-rigid tracking, co-authoring the widely used "Levenberg-Marquardt" algorithm tutorial and the influential "Bundler" structure-from-motion system. His work on the "Imagine" 3D vision system demonstrated early practical applications of range data for robotic assembly and navigation. With over 30,000 citations, his papers on robust model fitting and real-time camera tracking have become essential references. At Microsoft Research and later as a Vice President at Amazon, he led teams advancing computer vision for products like HoloLens and AWS Rekognition. Fitzgibbon's research has earned him multiple best paper awards and a lasting impact on both academic theory and industrial applications in 3D vision.