Data to be translated: Graduate students at Stanford University have developed an application called PIGEON that can determine specific locations with impressive accuracy by examining Google Street View images. PIGEON can predict the country of capture with 92% accuracy and, in 40% of guesses, can pinpoint the location within 25 kilometers of the target. The model is trained on a neural network called CLIP developed by OpenAI, as well as a dataset from the GeoGuessr game, achieving impressive results. The students also developed a separate model called PIGEOTTO, which, by training on 4 million photos from Flickr and Wikipedia, can identify locations from a single image.