Flicke-30K

The 'Flicke-30K' dataset contains images and their corresponding captions and is used to evaluate joint image-text retrieval methods.

Papers using Flicke-30K (1)

Learning to embed semantic similarity for joint image-text retrieval2022