Recognize Anything: A Strong Image Tagging Model

  • It looks like the comparison images at the top have inaccurate captions?