Geospatial Tracking

Geospatial tracking focuses on accurately determining and continuously monitoring the location of objects, a crucial task for applications ranging from autonomous vehicles to infrastructure management. Current research emphasizes robust methods for fusing data from diverse sensors, such as cameras and lidar, to improve accuracy and handle challenging conditions, often employing deep learning models for feature extraction and pose estimation within differentiable frameworks. This work is driven by the need for reliable localization beyond GPS, particularly in environments with limited or unreliable GNSS signals, and is leading to advancements in both algorithmic efficiency and the creation of large, publicly available datasets for benchmarking and model development.