README.md

metadata

dataset_info:
  features:
    - name: image
      dtype: image
    - name: text
      dtype: string
  splits:
    - name: train
      num_bytes: 333961906
      num_examples: 640
    - name: validation
      num_bytes: 91432668
      num_examples: 160
  download_size: 404977855
  dataset_size: 425394574
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: validation
        path: data/validation-*
task_categories:
  - image-to-text
  - visual-question-answering
language:
  - is
pretty_name: Icelandic Image Caption Dataset
size_categories:
  - n<1K

Icelandic Image Caption Dataset

curated by: Sigurdur Haukur Birgisson

The following data was collected from Wikipedia, using their api.

The dataset consist of an image and its respective description/caption, the description is in Icelandic.