GUESS THE LOCATION BY LOOKING AT THE PICTURE: STANFORD'S LATEST AI MODEL PIGEON HAS AN ACCURACY RATE OF OVER 90%

Bit — A team of researchers at Stanford University recently collaborated to develop an AI tool, PIGEON, that combines semantic geocell creation with label smoothing, pre-trained Street View images on CLIP vision transformers, and uses ProtoNets to refine location predictions on a set of candidate geographic units. PIGEON achieved a 91.96% accuracy rate and a 40.36% guess within 25 kilometers of the target on the “Guess the Country by Photo” subtask, which is the first state-of-the-art image geolocation paper in the past five years without military funding.

View Original
This page may contain third-party content, which is provided for information purposes only (not representations/warranties) and should not be considered as an endorsement of its views by Gate, nor as financial or professional advice. See Disclaimer for details.
  • Reward
  • Comment
  • Repost
  • Share
Comment
0/400
No comments
Trade Crypto Anywhere Anytime
qrCode
Scan to download Gate App
Community
English
  • 简体中文
  • English
  • Tiếng Việt
  • 繁體中文
  • Español
  • Русский
  • Français (Afrique)
  • Português (Portugal)
  • Bahasa Indonesia
  • 日本語
  • بالعربية
  • Українська
  • Português (Brasil)