TREC-AToMiC/AToMiC-Images-v0.2

Dataset Card for "AToMiC-All-Images_wi-pixels" Languages The dataset contains 108 languages in Wikipedia. Data Instances Each instance is an image, its representation in bytes, and its associated captions. Intended Usage Image collection for Text-to-Image retrieval Image--Caption Retrieval/Generation/Translation Licensing Information CC BY-SA 4.0 international license Citation Information TBA… See the full description on the dataset page: https://huggingface.co/datasets/TREC-AToMiC/AToMiC-Images-v0.2.