Recognize Anything: A Strong Image Tagging Model
(
4 comments
| on HN
)
Re:Recognize Anything: A Strong Image Tagging Model
by zooi on Sunday June 11, 2023 @11:06AM
(#36282069)
It looks like the comparison images at the top have inaccurate captions?
[ Parent
| on HN ]
Re:Recognize Anything: A Strong Image Tagging Model
by MitPitt on Monday June 12, 2023 @12:20AM
(#36288417)
RAM looks correct to me. What's wrong?
[ Parent
| on HN ]
Re:Recognize Anything: A Strong Image Tagging Model
by zooi on Wednesday June 14, 2023 @05:07AM
(#36323326)
E.g BLIP image is captioned "missing: dog" despite the dog being detected in the image.
[ Parent
| on HN ]
Re:Recognize Anything: A Strong Image Tagging Model
by MitPitt on Friday June 23, 2023 @10:45AM
(#36447515)
That's why it's a comparison. It's a page about RAM and how it's better than BLIP
[ Parent
| on HN ]