We propose KoSim-GL, a large-scale vision-based geo-localization dataset for drone positioning in GPS-denied environments. Geo-localization estimates a drone’s location by matching drone-view imagery against a geo-referenced satellite image database, offering a reliable alternative to GPS under conditions such as signal jamming, spoofing, or degradation in dense urban canyons. Although this task is challenging due to the domain gap between drone-view and satellite-view imagery, existing benchmarks are built predominantly around urban environments in the United States and China, leaving South Korea largely unrepresented, despite its distinctive landscape in which mountainous terrain coexists with dense high-rise districts and low-rise residential neighborhoods. To address this gap, we introduce KoSim-GL, constructed from drone-view images captured via an AirSim- and ROS-based flight simulator and satellite images collected through the Google Maps Tile API, covering the urban area of Daejeon, South Korea. Its key feature is a multi-view configuration that simultaneously captures five views, one nadir and four oblique, at each flight position across altitudes from 100 m to 600 m, enabling robust localization even in feature-sparse environments where nadir-only matching is prone to fail. In total, KoSim-GL comprises 2,450,315 drone images and 1704 satellite images. We further provide systematic comparisons against five existing benchmarks and baseline evaluations of ten representative geo-localization models under single- and multi-view settings. Experimental results show that the multi-view configuration substantially improves localization performance; for example, FSRA improves Recall@1 from 44.08% (single-view) to 65.37% (multi-view), a gain of 21.29 percentage points. The dataset is publicly available.

Building models are widely used as core spatial data in digital twins and urban simulations, where user-defined scenarios such as disasters or long-term material degradation must be visually reproduced. Yet the textures of these models are typically stored in fragmented atlas form and encode only the building appearance at the time of capture, making it difficult to adapt facade styles to diverse scenarios and often causing visual inconsistency when reused across different conditions. To address this limitation, we propose a facade-based texture style transfer pipeline that reconstructs the texture atlas into facade-aligned images where architectural components are clearly represented, and then performs style transfer while preserving the shapes and relative positions of key elements such as windows and walls. Our method builds on Stable Diffusion XL, using window-wall masks and edge-based structural conditioning to maintain geometry, and IP-Adapter to control color and material style from user-defined reference images. Experiments on 3,370 facade images generated from 886 real building models under three degradation scenarios show that the proposed approach achieves a more balanced trade-off between style faithfulness and structural preservation than recent diffusion-based baselines, while qualitatively retaining the geometric consistency of essential architectural components.