Crowd-sourced captions describing the picture. Multiple captions per picture (total = 158k), stored in a seperate csv file together with image ids and names.
Images are from variety of themes ('everyday activities, events and scenes') from Flickr, photograph sharing site for both amateurs and professionals.
~30k images
Original paper: From image descriptions to visual denotations: New similarity metrics for semantic inference over event descriptions
Homepage:… See the full description on the dataset page:
https://huggingface.co/datasets/TPagent/Flickr30k.
