Abstract

Recent works show that it is possible to train a deep network to determine the geographic location of a ground-level image (e.g., a Google street-view panorama) by matching it against a satellite map covering the wide geographic area of interest. Conventional deep networks, which often cast the problem as a metric embedding task, however, suffer from poor performance in terms of low recall rates. One of the key reasons is the vast differences between the two view modalities, i.e., ground view versus aerial/satellite view. They not only exhibit very different visual appearances, but also have distinctive geometric geometric configurations. Existing deep methods overlook those appearance and geometric differences, and instead use a bruteforce training procedure, leading to inferior performance. In this paper, we develop a new deep network to explicitly address these inherent differences between ground and aerial views. We observe that pixels lying on the same azimuth direction in an aerial image approximately correspond to a vertical image column in the ground view image. Thus, we propose a two-step approach to exploit this prior. The first step is to apply a regular polar transform to warp an aerial image such that its domain is closer to that of a ground-view panorama. Note that polar transform as a pure geometric transformation is agnostic to scene content, hence cannot bring the two domains into full alignment. Then, we add a subsequent spatial-attention mechanism which brings corresponding deep features closer in the embedding space. To improve the robustness of feature representation, we introduce a feature aggregation strategy via learning multiple spatial embeddings. By the above two-step approach, we achieve more discriminative deep representations, facilitating cross-view Geo-localization more accurate. Our experiments on standard benchmark datasets show significant performance boosting, achieving more than doubled recall rate compared with the previous state of the art. Remarkably, the recall rate@top-1 improves from 22.5% in [4] (or 40.7% in [10]) to 89.8% on CVUSA benchmark, and from 20.1% [4] to 81.0% on the new CVACT dataset.