CLIP, BLIP, or a combination of them?
https://huggingface.co/spaces/pharma/CLIP-Interrogator
But it doesn't describe every object in the image
https://huggingface.co/spaces/pharma/CLIP-Interrogator
But it doesn't describe every object in the image
Is there an AI that can accurately label pictures? The examples I've seen so far can recognize a few animals and objects and that's it.
There are people in that photo, sure, but are they wearing a jacket or a bikini? Are they happy or sad? Which model and make is that car?
Google Photos seems to do a better job at this in general but doesn't expose the tags, you can only search them.