Recognize Anything: A Strong Image Tagging Model (recognize-anything.github.io) 18 points by MitPitt 3y ago ↗ HN
[–] zooi 3y ago ↗ It looks like the comparison images at the top have inaccurate captions? [–] MitPitt 3y ago ↗ RAM looks correct to me. What's wrong? [–] zooi 3y ago ↗ E.g BLIP image is captioned "missing: dog" despite the dog being detected in the image. [–] MitPitt 3y ago ↗ That's why it's a comparison. It's a page about RAM and how it's better than BLIP
[–] MitPitt 3y ago ↗ RAM looks correct to me. What's wrong? [–] zooi 3y ago ↗ E.g BLIP image is captioned "missing: dog" despite the dog being detected in the image. [–] MitPitt 3y ago ↗ That's why it's a comparison. It's a page about RAM and how it's better than BLIP
[–] zooi 3y ago ↗ E.g BLIP image is captioned "missing: dog" despite the dog being detected in the image. [–] MitPitt 3y ago ↗ That's why it's a comparison. It's a page about RAM and how it's better than BLIP
[–] MitPitt 3y ago ↗ That's why it's a comparison. It's a page about RAM and how it's better than BLIP
4 comments
[ 3.6 ms ] story [ 18.2 ms ] thread