Anonymous Captions Dataset is a collection of
{image, generated anonymous caption?Anonymous captions are image captions that do not refer to specific visible objects but instead capture the relational logic conveyed by the image.} pairs, split into Train (100k) and Test (14k) sets. You are currently viewing the Test Split.
All images are extracted from the publicly available
LAION dataset. Images are shown by their original Image URL. Some images may not be available because their original URLs are no longer active. We do not own any rights to the images and acknowledge the original authors of each images.
Want to download the dataset or access the captioning model? Click on this Github Repo.