metadata

language:
  - en
  - zh
license: apache-2.0
configs:
  - config_name: default
    data_files:
      - split: validation
        path: data/validation-*
      - split: test
        path: data/test-*
      - split: train
        path: data/train-*
dataset_info:
  features:
    - name: ref_list
      list:
        - name: bbox
          sequence: float64
        - name: category
          dtype: string
        - name: dialog
          sequence:
            sequence: string
        - name: dialog_cn
          sequence:
            sequence: string
        - name: id
          dtype: string
    - name: image_info
      struct:
        - name: file_name
          dtype: string
        - name: height
          dtype: int64
        - name: id
          dtype: string
        - name: width
          dtype: int64
    - name: image
      dtype: image
  splits:
    - name: validation
      num_bytes: 96380848
      num_examples: 996
    - name: test
      num_bytes: 193325330.698
      num_examples: 1997
    - name: train
      num_bytes: 1735786813.55
      num_examples: 17710
  download_size: 865015922
  dataset_size: 2025492992.248

Dataset Card for "invig"

Github

@misc{invigdataset,
    title={InViG: Interactive Visual-Language Disambiguation with 21K Human-to-Human Dialogues},
    author={Zhang, Hanbo and Mo, Yuchen and Xu, Jie and Si, Qingyi and Kong, Tao},
    howpublished = {\url{https://github.com/ZhangHanbo/invig-dataset}},
    year={2023}
}