PIGEON: Predicting Image Geolocations

• URL: http://arxiv.org/abs/2307.05845v6
• Date: Tue, 28 May 2024 22:00:10 GMT
• Title: PIGEON: Predicting Image Geolocations
• Authors: Lukas Haas, Michal Skreta, Silas Alberti, Chelsea Finn,
• Abstract summary: We present a new geolocalization system that combines semantic geocell creation, multi-task contrastive pretraining, and a novel loss function. PIGEOTTO is the first image geolocalization model that effectively generalizes to unseen places.
• Score: 44.99833362998488
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Planet-scale image geolocalization remains a challenging problem due to the diversity of images originating from anywhere in the world. Although approaches based on vision transformers have made significant progress in geolocalization accuracy, success in prior literature is constrained to narrow distributions of images of landmarks, and performance has not generalized to unseen places. We present a new geolocalization system that combines semantic geocell creation, multi-task contrastive pretraining, and a novel loss function. Additionally, our work is the first to perform retrieval over location clusters for guess refinements. We train two models for evaluations on street-level data and general-purpose image geolocalization; the first model, PIGEON, is trained on data from the game of Geoguessr and is capable of placing over 40% of its guesses within 25 kilometers of the target location globally. We also develop a bot and deploy PIGEON in a blind experiment against humans, ranking in the top 0.01% of players. We further challenge one of the world's foremost professional Geoguessr players to a series of six matches with millions of viewers, winning all six games. Our second model, PIGEOTTO, differs in that it is trained on a dataset of images from Flickr and Wikipedia, achieving state-of-the-art results on a wide range of image geolocalization benchmarks, outperforming the previous SOTA by up to 7.7 percentage points on the city accuracy level and up to 38.8 percentage points on the country level. Our findings suggest that PIGEOTTO is the first image geolocalization model that effectively generalizes to unseen places and that our approach can pave the way for highly accurate, planet-scale image geolocalization systems. Our code is available on GitHub.

Related papers

• CityGuessr: City-Level Video Geo-Localization on a Global Scale [54.371452373726584]
  We propose a novel problem of worldwide video geolocalization with the objective of hierarchically predicting the correct city, state/province, country, and continent, given a video. No large scale video datasets that have extensive worldwide coverage exist, to train models for solving this problem. We introduce a new dataset, CityGuessr68k comprising of 68,269 videos from 166 cities all over the world.
  arXiv  Detail & Related papers  (2024-11-10T03:20:00Z)
• G3: An Effective and Adaptive Framework for Worldwide Geolocalization Using Large Multi-Modality Models [40.69217368870192]
  We propose a novel framework for worldwide geolocalization based on Retrieval-Augmented Generation (RAG) G3 consists of three steps, i.e., Geo-alignment, Geo-diversification, and Geo-verification. Experiments on two well-established datasets verify the superiority of G3 compared to other state-of-the-art methods.
  arXiv  Detail & Related papers  (2024-05-23T15:37:06Z)
• GeoCLIP: Clip-Inspired Alignment between Locations and Images for Effective Worldwide Geo-localization [61.10806364001535]
  Worldwide Geo-localization aims to pinpoint the precise location of images taken anywhere on Earth. Existing approaches divide the globe into discrete geographic cells, transforming the problem into a classification task. We propose GeoCLIP, a novel CLIP-inspired Image-to-GPS retrieval approach that enforces alignment between the image and its corresponding GPS locations.
  arXiv  Detail & Related papers  (2023-09-27T20:54:56Z)
• Where We Are and What We're Looking At: Query Based Worldwide Image Geo-localization Using Hierarchies and Scenes [53.53712888703834]
  We introduce an end-to-end transformer-based architecture that exploits the relationship between different geographic levels. We achieve state of the art street level accuracy on 4 standard geo-localization datasets.
  arXiv  Detail & Related papers  (2023-03-07T21:47:58Z)
• G^3: Geolocation via Guidebook Grounding [92.46774241823562]
  We study explicit knowledge from human-written guidebooks that describe the salient and class-discriminative visual features humans use for geolocation. We propose the task of Geolocation via Guidebook Grounding that uses a dataset of StreetView images from a diverse set of locations. Our approach substantially outperforms a state-of-the-art image-only geolocation method, with an improvement of over 5% in Top-1 accuracy.
  arXiv  Detail & Related papers  (2022-11-28T16:34:40Z)
• Where in the World is this Image? Transformer-based Geo-localization in the Wild [48.69031054573838]
  Predicting the geographic location (geo-localization) from a single ground-level RGB image taken anywhere in the world is a very challenging problem. We propose TransLocator, a unified dual-branch transformer network that attends to tiny details over the entire image. We evaluate TransLocator on four benchmark datasets - Im2GPS, Im2GPS3k, YFCC4k, YFCC26k and obtain 5.5%, 14.1%, 4.9%, 9.9% continent-level accuracy improvement.
  arXiv  Detail & Related papers  (2022-04-29T03:27:23Z)
• Rethinking Visual Geo-localization for Large-Scale Applications [18.09618985653891]
  We build San Francisco eXtra Large, a new dataset covering a whole city and providing a wide range of challenging cases. We design a new highly scalable training technique, called CosPlace, which casts the training as a classification problem. We achieve state-of-the-art performance on a wide range of datasets and find that CosPlace is robust to heavy domain changes.
  arXiv  Detail & Related papers  (2022-04-05T15:33:45Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.