TasksSotADatasetsPapersMethodsSubmitAbout
Papers With Code 2

A community resource for machine learning research: papers, code, benchmarks, and state-of-the-art results.

Explore

Notable BenchmarksAll SotADatasetsPapersMethods

Community

Submit ResultsAbout

Data sourced from the PWC Archive (CC-BY-SA 4.0). Built by the community, for the community.

Models/GeoCLIP

GeoCLIP

Reported on 32 benchmarks across 3 tasks · 1 paper · 6 SOTA

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Computer Vision15 results

  • Image ClassificationonIm2GPS3k
    Street level (1 km)· 2023-09-27
    14.1
    best: 18.79 (GeoRanker)
    SOTA
    GeoCLIP: Clip-Inspired Alignment between Locations and Images for Effective Worldwide Geo-localizationarXiv:2309.16020
  • Image ClassificationonYFCC26k
    Street level (1 km)· 2023-09-27
    11.6
    SOTA
    GeoCLIP: Clip-Inspired Alignment between Locations and Images for Effective Worldwide Geo-localizationarXiv:2309.16020
  • Image ClassificationonIm2GPS3k
    City level (25 km)· 2023-09-27
    34.5
    best: 45.05 (GeoRanker)
    GeoCLIP: Clip-Inspired Alignment between Locations and Images for Effective Worldwide Geo-localizationarXiv:2309.16020
  • Image ClassificationonIm2GPS3k
    Continent level (2500 km)· 2023-09-27
    83.8
    best: 89.29 (GeoRanker)
    GeoCLIP: Clip-Inspired Alignment between Locations and Images for Effective Worldwide Geo-localizationarXiv:2309.16020
  • Image ClassificationonIm2GPS3k
    Country level (750 km)· 2023-09-27
    69.7
    best: 76.31 (GeoRanker)
    GeoCLIP: Clip-Inspired Alignment between Locations and Images for Effective Worldwide Geo-localizationarXiv:2309.16020
  • Image ClassificationonIm2GPS3k
    Region level (200 km)· 2023-09-27
    50.7
    best: 61.49 (GeoRanker)
    GeoCLIP: Clip-Inspired Alignment between Locations and Images for Effective Worldwide Geo-localizationarXiv:2309.16020
  • Image ClassificationonYFCC26k
    City level (25 km)· 2023-09-27
    22.2
    best: 25.8 (PIGEOTTO)
    GeoCLIP: Clip-Inspired Alignment between Locations and Images for Effective Worldwide Geo-localizationarXiv:2309.16020
  • Image ClassificationonYFCC26k
    Continent level (2500 km)· 2023-09-27
    76
    best: 79 (PIGEOTTO)
    GeoCLIP: Clip-Inspired Alignment between Locations and Images for Effective Worldwide Geo-localizationarXiv:2309.16020
  • Image ClassificationonYFCC26k
    Country level (750 km)· 2023-09-27
    57.5
    best: 63.2 (PIGEOTTO)
    GeoCLIP: Clip-Inspired Alignment between Locations and Images for Effective Worldwide Geo-localizationarXiv:2309.16020
  • Image ClassificationonYFCC26k
    Region level (200 km)· 2023-09-27
    36.7
    best: 42.7 (PIGEOTTO)
    GeoCLIP: Clip-Inspired Alignment between Locations and Images for Effective Worldwide Geo-localizationarXiv:2309.16020
  • Image ClassificationonGWS15k
    City level (25 km)· 2023-09-27
    3.1
    best: 9.2 (PIGEOTTO)
    GeoCLIP: Clip-Inspired Alignment between Locations and Images for Effective Worldwide Geo-localizationarXiv:2309.16020
  • Image ClassificationonGWS15k
    Continent level (2500 km)· 2023-09-27
    74.1
    best: 85.1 (PIGEOTTO)
    GeoCLIP: Clip-Inspired Alignment between Locations and Images for Effective Worldwide Geo-localizationarXiv:2309.16020
  • Image ClassificationonGWS15k
    Country level (750 km)· 2023-09-27
    45.7
    best: 65.7 (PIGEOTTO)
    GeoCLIP: Clip-Inspired Alignment between Locations and Images for Effective Worldwide Geo-localizationarXiv:2309.16020
  • Image ClassificationonGWS15k
    Region level (200 km)· 2023-09-27
    16.9
    best: 31.2 (PIGEOTTO)
    GeoCLIP: Clip-Inspired Alignment between Locations and Images for Effective Worldwide Geo-localizationarXiv:2309.16020
  • Image ClassificationonGWS15k
    Street level (1 km)· 2023-09-27
    0.6
    best: 0.7 (PIGEOTTO)
    GeoCLIP: Clip-Inspired Alignment between Locations and Images for Effective Worldwide Geo-localizationarXiv:2309.16020

Graphs15 results

  • 4K 60FpsonIm2GPS3k
    Street level (1 km)· 2023-09-27
    14.1
    best: 18.79 (GeoRanker)
    SOTA
    GeoCLIP: Clip-Inspired Alignment between Locations and Images for Effective Worldwide Geo-localizationarXiv:2309.16020
  • 4K 60FpsonYFCC26k
    Street level (1 km)· 2023-09-27
    11.6
    SOTA
    GeoCLIP: Clip-Inspired Alignment between Locations and Images for Effective Worldwide Geo-localizationarXiv:2309.16020
  • 4K 60FpsonIm2GPS3k
    City level (25 km)· 2023-09-27
    34.5
    best: 45.05 (GeoRanker)
    GeoCLIP: Clip-Inspired Alignment between Locations and Images for Effective Worldwide Geo-localizationarXiv:2309.16020
  • 4K 60FpsonIm2GPS3k
    Continent level (2500 km)· 2023-09-27
    83.8
    best: 89.29 (GeoRanker)
    GeoCLIP: Clip-Inspired Alignment between Locations and Images for Effective Worldwide Geo-localizationarXiv:2309.16020
  • 4K 60FpsonIm2GPS3k
    Country level (750 km)· 2023-09-27
    69.7
    best: 76.31 (GeoRanker)
    GeoCLIP: Clip-Inspired Alignment between Locations and Images for Effective Worldwide Geo-localizationarXiv:2309.16020
  • 4K 60FpsonIm2GPS3k
    Region level (200 km)· 2023-09-27
    50.7
    best: 61.49 (GeoRanker)
    GeoCLIP: Clip-Inspired Alignment between Locations and Images for Effective Worldwide Geo-localizationarXiv:2309.16020
  • 4K 60FpsonYFCC26k
    City level (25 km)· 2023-09-27
    22.2
    best: 25.8 (PIGEOTTO)
    GeoCLIP: Clip-Inspired Alignment between Locations and Images for Effective Worldwide Geo-localizationarXiv:2309.16020
  • 4K 60FpsonYFCC26k
    Continent level (2500 km)· 2023-09-27
    76
    best: 79 (PIGEOTTO)
    GeoCLIP: Clip-Inspired Alignment between Locations and Images for Effective Worldwide Geo-localizationarXiv:2309.16020
  • 4K 60FpsonYFCC26k
    Country level (750 km)· 2023-09-27
    57.5
    best: 63.2 (PIGEOTTO)
    GeoCLIP: Clip-Inspired Alignment between Locations and Images for Effective Worldwide Geo-localizationarXiv:2309.16020
  • 4K 60FpsonYFCC26k
    Region level (200 km)· 2023-09-27
    36.7
    best: 42.7 (PIGEOTTO)
    GeoCLIP: Clip-Inspired Alignment between Locations and Images for Effective Worldwide Geo-localizationarXiv:2309.16020
  • 4K 60FpsonGWS15k
    City level (25 km)· 2023-09-27
    3.1
    best: 9.2 (PIGEOTTO)
    GeoCLIP: Clip-Inspired Alignment between Locations and Images for Effective Worldwide Geo-localizationarXiv:2309.16020
  • 4K 60FpsonGWS15k
    Continent level (2500 km)· 2023-09-27
    74.1
    best: 85.1 (PIGEOTTO)
    GeoCLIP: Clip-Inspired Alignment between Locations and Images for Effective Worldwide Geo-localizationarXiv:2309.16020
  • 4K 60FpsonGWS15k
    Country level (750 km)· 2023-09-27
    45.7
    best: 65.7 (PIGEOTTO)
    GeoCLIP: Clip-Inspired Alignment between Locations and Images for Effective Worldwide Geo-localizationarXiv:2309.16020
  • 4K 60FpsonGWS15k
    Region level (200 km)· 2023-09-27
    16.9
    best: 31.2 (PIGEOTTO)
    GeoCLIP: Clip-Inspired Alignment between Locations and Images for Effective Worldwide Geo-localizationarXiv:2309.16020
  • 4K 60FpsonGWS15k
    Street level (1 km)· 2023-09-27
    0.6
    best: 0.7 (PIGEOTTO)
    GeoCLIP: Clip-Inspired Alignment between Locations and Images for Effective Worldwide Geo-localizationarXiv:2309.16020

Miscellaneous2 results

  • GPS EmbeddingsonGeo-Tagged NUS-WIDE (GPS + Visual)
    mAP· 2023-09-27
    0.362
    SOTA
    GeoCLIP: Clip-Inspired Alignment between Locations and Images for Effective Worldwide Geo-localizationarXiv:2309.16020
  • GPS EmbeddingsonGeo-Tagged NUS-WIDE (GPS Only)
    mAP· 2023-09-27
    0.249
    SOTA
    GeoCLIP: Clip-Inspired Alignment between Locations and Images for Effective Worldwide Geo-localizationarXiv:2309.16020