Visual Storytelling Dataset (VIST)

The Visual Storytelling Dataset (VIST) consists of 10,117 Flickr albums and 210,819 unique images. Each sample is one sequence of 5 photos selected from the same album paired with a single human constructed story, where each story is comprised of mostly one sentence per image.

Data and Resources

Cite this as

Junjie Hu, Yu Cheng, Zhe Gan, Jingjing Liu, Jianfeng Gao, Graham Neubig (2024). Dataset: Visual Storytelling Dataset (VIST). https://doi.org/10.57702/ucqupobe

DOI retrieved: December 16, 2024

Additional Info

Field Value
Created December 16, 2024
Last update December 16, 2024
Defined In https://doi.org/10.48550/arXiv.2205.03774
Citation
  • https://doi.org/10.48550/arXiv.1909.05316
Author Junjie Hu
More Authors
Yu Cheng
Zhe Gan
Jingjing Liu
Jianfeng Gao
Graham Neubig
Homepage https://github.com/JunjieHu/ReCo-RL