Below is a list of datasets that we have made publicly available for research purposes.

• Archive of Many Outdoor Scenes (AMOS): A dataset containing hundreds of millions of images gathered from publicly available outdoor webcams.
• Brooklyn and Queens (Near/Remote): A large dataset containing millions of pairs of ground-level and aerial/satellite images from across the United States.
• Crossview Sound (CVS): A large dataset containing many pairs of geotagged audio files and aerial/satellite images.
• Crossview USA (CVUSA): A large dataset containing millions of pairs of ground-level and aerial/satellite images from across the United States.
• Face2Year: A large number of images extracted from highschool yearbooks.
• GeoFaces: A dataset containing millions of geotagged images of human faces.
• Horizon Lines in the Wild: A large dataset of images annotated with horizon lines.
• LOST: A large dataset containing streaming video collected from publicly available outdoor webcams. The project website no longer exists, but we still have the data. Please reach out if you are interested.
• SkyFinder: A large dataset of webcam images annotated with sky regions.
• Timelapse Clouds: A dataset containing videos of clouds and cloud shadows captured by stationary ground-based cameras.