Get Image Description using Joy Caption
Theres this thing on huggingface that allows me to put in an image and then ask it to des
cribe it:
https://huggingface.co/spaces/fancyfeast/joy-caption-alpha-two
Its really good. Would be great to be able to upload image and then ask it to identify it.
User could change settings how they want the identify it as seen in the above linked demo.
Log in to comment and vote
Comments3
Lime Koala
May 20, 2025
Venice having this on API would undoubtedly provide significant benefits to the open source model training community down the line. It doesn’t need a ton of VRAM and the performance is excellent.
Plum Ash
Feb 14, 2025
Would love to have this added, this way we can generate and export images into text models or code models to improve workflow.
Azure Leaf
Jan 23, 2025
i tequest too that ceature it’s for immersion purposes in roleplay or simply have a fun reaction of my ai. 👍