Swahili Image Captioning Dataset
Grevians Onula, Shikali Casper, Omala Andrew · Zenodo (CERN European Organization for Nuclear Research) · 2023
The SwaFlickr8k dataset is an extension of the well-known Flickr8k dataset, specifically designed for image captioning tasks. It includes a collection of images and corresponding captions written in Swahili. With 8,091 unique images and 40,455 captions, this dataset provides a valuable resource for research and development in the field of image understanding and language processing, particularly in the context of the Swahili language.