Abstract: Grounding language to the visual observations of a navigating agent can be performed using off-the-shelf visual-language models pretrained on Internet-scale data (e.g., image captions).

Abstract: This article investigates loosely coupled inertial navigation system/global positioning system (INS/GPS) integration for land vehicle navigation. To achieve navigation with higher accuracy ...