Chair: Maani Ghaffari

Abstract:
Robot localization uses external sensor data to estimate a robot's position, but movements can alter vision measurements and learned features, impacting robustness. This thesis introduces a new perspective on multi-modal robot localization through the integration of geometric learning, with a primary focus on equivariant networks to address these challenges. First, we explore the application of geometric learning in the loop closure process, enabling the robust identification of previously visited locations despite transformative changes. Next, we propose SE(3)-equivariant transformer designs for learning point cloud correspondences, crucial for point cloud registration tasks. Finally, we integrate geometric learning with visual foundation models. Overall, this dissertation highlights the advantages of incorporating geometric constraints into machine learning models, significantly enhancing both the reliability and performance of robot localization and contributing to critical advancements in robotics.

Zoom link: https://umich.zoom.us/j/91678338359?pwd=cLzyN6l1MU1exRFE25hKzb2uW9EMmW.1