README.md

metadata

language:
  - ko
  - en
size_categories:
  - 10B<n<100B
task_categories:
  - visual-question-answering
dataset_info:
  features:
    - name: id
      dtype: string
    - name: image
      dtype: image
    - name: caption
      dtype: string
    - name: caption_ls
      list: string
    - name: category
      dtype: string
    - name: objects
      list:
        - name: id
          dtype: int32
        - name: text
          dtype: string
        - name: bbox
          list: int32
    - name: question_answer
      list:
        - name: id
          dtype: string
        - name: question
          dtype: string
        - name: answer
          dtype: string
        - name: visual_concept
          list: string
        - name: question_en
          dtype: string
        - name: answer_en
          dtype: string
        - name: question_type
          list: string
        - name: ans_source
          dtype: string
        - name: kb_source
          list: string
        - name: fact
          list: string
    - name: metadata
      struct:
        - name: KOREAN_IMAGE
          dtype: string
        - name: DIRECT_IMAGE
          dtype: string
        - name: ACTION
          dtype: string
        - name: IMAGE_ACQUISITION_DATE
          dtype: string
        - name: IMAGE_URL
          dtype: string
        - name: LICENSE
          dtype: string
        - name: MAINOBJECT
          dtype: string
        - name: MAINOBJECT_URL
          dtype: string
        - name: SCENE
          dtype: string
        - name: SUBOBJECT_1
          dtype: string
        - name: SUBOBJECT_2
          dtype: string
        - name: WIDTH
          dtype: string
        - name: HEIGHT
          dtype: string
        - name: WIKI_FILE
          dtype: string
  splits:
    - name: train
      num_bytes: 11169890399.625
      num_examples: 47203
    - name: validation
      num_bytes: 1548226868.125
      num_examples: 6423
  download_size: 12644146792
  dataset_size: 12718117267.75
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: validation
        path: data/validation-*
tags:
  - Caption
  - Image