search

cwj1412 / MSCOCO-Flikcr30K_FG

Benchmark data for "Rethinking Benchmarks for Cross-modal Image-text Retrieval" (SIGIR 2023)
22 stars 0 forks source link

text captions #4

Open frozen30 opened 11 months ago

frozen30 commented 11 months ago

Hello, dear author, after perusing your article, I have a question I'd like to pose: From where do the text captions for extended images, apart from those in the MSCOCO 5k and Flickr 1k datasets, originate?