Introduction Remarks

https://huggingface.co/spaces/pharma/CLIP-Interrogator

Put an image, and it will generate a CLIP prompt. The CLIP prompt is intended to be supplied into

-- hand written notes --

Looking at the Hugging Face App code:

https://huggingface.co/spaces/pharma/CLIP-Interrogator/blob/main/app.py

Note the reference data: https://huggingface.co/spaces/pharma/CLIP-Interrogator/tree/main/data

The app will mix and match the various phrases to generate captions based on the uploaded image

https://huggingface.co/spaces/pharma/CLIP-Interrogator/blob/main/data/artists.txt

A. B. Jackson
A. J. Casson
A. R. Middleton Todd
A.B. Frost
A.D.M. Cooper
Aaron Bohrod
Aaron Douglas
Aaron Jasinski
Aaron Miller
Aaron Nagel

BLIP Language Model

Bootstrapping Language Image Pre-training for Unified Vision-Language understanding

arxiv paper

This is specifically designed to give an ok caption from an image, but it is NOT the inverse of the CLIP encoder