My work is in the intersection of 3D computer vision and machine learning, with the goal of making 3D computer vision algorithms such as 3D reconstruction and visual localization more robust and reliable through scene understanding while using 3D computer vision methods to train machine learning models.

My research interests include (large-scale) image-based localization, real-time localization, 3D mapping and reconstruction, implicit neural scene representations and neural rendering, learning local features, machine learning for 3D computer vision, and camera calibration, with applications in self-driving cars and other autonomous robots as well as augmented and virtual reality. My current work focuses on making algorithms for localization and mapping "smarter" (and thus more robust and reliable) by incorporating higher-level scene understanding into the process.