Flickr30k

Flickr30k is a dataset that contains images paired with descriptive captions, used to evaluate multimodal models on tasks such as image captioning and cross-modal retrieval.

Papers using Flickr30k (3)

Visualsparta: An Embarrassingly Simple Approach To Large-scale Text-to-image Search With Weighted Bag-of-words2021 · 22 citesThinking Fast and Slow: Efficient Text-to-Visual Retrieval with Transformers2021 · 133 citesNearest Neighbor Normalization Improves Multimodal Retrieval2024