Golden Dataset ‐ Categories - AIMLOps-C4-G16/aimlops-capstone-project GitHub Wiki

We will create image databases with ~5 captions per image for additional fine-tuning of our image captioning model as per our subjective expectations from it.

These are the current list of categories that we want to recognize:

  • Indian monuments - 25 monuments * 10 images * 5 captions
  • Indian food - 50 foods * 10 images * 5 captions
  • Indian festivals - 20 festivals * 10 images * 5 captions
  • Indian sports
  • Indian state costumes (link)