[2603.25686] Just Zoom In: Cross-View Geo-Localization via Autoregressive Zooming

Computer Science > Computer Vision and Pattern Recognition arXiv:2603.25686 (cs) [Submitted on 26 Mar 2026] Title:Just Zoom In: Cross-View Geo-Localization via Autoregressive Zooming Authors:Yunus Talha Erzurumlu, Jiyong Kwag, Alper Yilmaz View a PDF of the paper titled Just Zoom In: Cross-View Geo-Localization via Autoregressive Zooming, by Yunus Talha Erzurumlu and 2 other authors View PDF HTML (experimental) Abstract:Cross-view geo-localization (CVGL) estimates a camera's location by matching a street-view image to geo-referenced overhead imagery, enabling GPS-denied localization and navigation. Existing methods almost universally formulate CVGL as an image-retrieval problem in a contrastively trained embedding space. This ties performance to large batches and hard negative mining, and it ignores both the geometric structure of maps and the coverage mismatch between street-view and overhead imagery. In particular, salient landmarks visible from the street view can fall outside a fixed satellite crop, making retrieval targets ambiguous and limiting explicit spatial inference over the map. We propose Just Zoom In, an alternative formulation that performs CVGL via autoregressive zooming over a city-scale overhead map. Starting from a coarse satellite view, the model takes a short sequence of zoom-in decisions to select a terminal satellite cell at a target resolution, without contrastive losses or hard negative mining. We further introduce a realistic benchmark with crowd-...