Dataset Groups Activity Stream High Quality Image-Text Pairs (HQITP) High Quality Image-Text Pairs (HQITP) dataset contains 134M high-quality image-caption pairs. BibTex: @dataset{Yuhui_Zhang_and_Brandon_McKinzie_and_Zhe_Gan_and_Vaishaal_Shankar_and_Alexander_Toshev_2024, abstract = {High Quality Image-Text Pairs (HQITP) dataset contains 134M high-quality image-caption pairs.}, author = {Yuhui Zhang and Brandon McKinzie and Zhe Gan and Vaishaal Shankar and Alexander Toshev}, doi = {10.57702/x0qiuh4s}, institution = {No Organization}, keyword = {'image captioning', 'image-text pairs'}, month = {dec}, publisher = {TIB}, title = {High Quality Image-Text Pairs (HQITP)}, url = {https://service.tib.eu/ldmservice/dataset/high-quality-image-text-pairs--hqitp-}, year = {2024} }