Shallow models meet deep vision

This project is concerned with perceiving, understanding, and representing objects and their environment. This constitutes key challenges for any autonomous system or augmented reality setup, especially when these functions are performed under uncertainty. The project combines statistical machine learning and computer vision to develop new methods for data-driven 3D visual computing, uncertainty quantification, and online learning. We put special interest in advancing the state-of-the-art in vision and sensor fusion methods, and in developing methods capable of inferring location, pose, and semantics from visual data under uncertainty. This is a crossover of the fields of computer vision and statistical machine learning, and aim at renewing the view on how the two can be combined.