The goal of image localization is to estimate the geographic location (or scene-relative location) of an image (or video). We have worked on this problem for over 15 years, from a variety of perspectives. Our early work focused on localizing static outdoor cameras using temporal variations as a cue. More recent work has focused on the use of deep neural networks for single image localization, with a significant focus on exploiting overhead imagery for ground-level image localization. Current research includes unconstrained cross-view pose estimation, goal modality agnostic active geo-localization (GOMAA-Geo), and localizing limited field-of-view images using cross-view matching (ArcGeo). We also develop structure-aware methods for direct pose estimation and extend absolute pose regression to multiple scenes.