Improving Multimodal Datasets with Image Captioning Paper • 2307.10350 • Published Jul 19, 2023 • 11