Image-Based Localization

Contains data from three platforms, i.e., synthetic drones, satellites and ground cameras of 1,652 university buildings around the world. University-1652 is a drone-based geo-localization dataset and enables two new tasks, i.e., drone-view target localization and drone navigation. Source: University-1652: A Multi-view Multi-source Benchmark for Drone-based Geo-localization

Similar to CVUSA and CVACT, the VIGOR dataset contains satellites and street imagery to match them to each other to find the location of the street imagery. For this purpose, data from 4 major American cities were used, namely San Francisco, New York, Seattle and Chicago. Unlike the previous datasets, there are two settings: The SAME-Area setting where images of all cities are available in training and validation split. Secondly, there is the CROSS area setting where training is done on two cities (New York, Seattle) and evaluation is done on Chicago and San Francisco. In addition, the dataset contains semi-positive images which are very close to an actual ground truth image and thus serve as a distraction for the matching task. In total, the dataset consists of 90,618 satellite images and 105,214 street images.

Cross-View Geo-Localisation within urban regions is challenging in part due to the lack of geo-spatial structuring within current datasets and techniques. We propose utilising graph representations to model sequences of local observations and the connectivity of the target location. Modelling as a graph enables generating previously unseen sequences by sampling with new parameter configurations. SpaGBOL contains 98,855 panoramic streetview images across different seasons, and 19,771 corresponding satellite images from 10 mostly densely populated international cities. This translates to 5 panoramic images and one satellite image per graph node. Downloading instructions below.

The appearance of the world varies dramatically not only from place to place but also from hour to hour and month to month. Every day billions of images capture this complex relationship, many of which are associated with precise time and location metadata. We propose to use these images to construct a global-scale, dynamic map of visual appearance attributes. Such a map enables fine-grained understanding of the expected appearance at any geographic location and time. Our approach integrates dense overhead imagery with location and time metadata into a general framework capable of mapping a wide variety of visual attributes. A key feature of our approach is that it requires no manual data annotation. We demonstrate how this approach can support various applications, including image-driven mapping, image geolocalization, and metadata verification.

The standard evaluation protocol of Cross-View Time dataset allows for certain cameras to be shared between training and testing sets. This protocol can emulate scenarios in which we need to verify the authenticity of images from a particular set of devices and locations. Considering the ubiquity of surveillance systems (CCTV) nowadays, this is a common scenario, especially for big cities and high visibility events (e.g., protests, musical concerts, terrorist attempts, sports events). In such cases, we can leverage the availability of historical photographs of that device and collect additional images from previous days, months, and years. This would allow the model to better capture the particularities of how time influences the appearance of that specific place, probably leading to a better verification accuracy. However, there might be cases in which data is originated from heterogeneous sources, such as social media. In this sense, it is essential that models are optimized on camera-disjoint sets to avoid learning sensor-specific characteristics that might not generalize accordingly for new imagery during inference. With this in mind, we propose a novel organization for CVT dataset. We split available data into training and testing sets, ensuring that all images from a single camera are assigned to the same set. During this division, we aimed to keep the size of each set roughly similar to the original splits, allowing models to be optimized with similar amounts of data.

The NAVER LABS localization datasets are 5 new indoor datasets for visual localization in challenging real-world environments. They were captured in a large shopping mall and a large metro station in Seoul, South Korea, using a dedicated mapping platform consisting of 10 cameras and 2 laser scanners. In order to obtain accurate ground truth camera poses, we used a robust LiDAR SLAM which provides initial poses that are then refined using a novel structure-from-motion based optimization. The datasets are provided in the kapture format and contain about 130k images as well as 6DoF camera poses for training and validation. We also provide sparse Lidar-based depth maps for the training images. The poses of the test set are withheld to not bias the benchmark.