metadata
dataset_info:
  features:
    - name: image
      dtype: image
    - name: image_id
      dtype: string
    - name: image_w_bbox
      dtype: image
    - name: ui_neg_objs_type_id
      sequence: int64
    - name: ui_neg_objs_clickable
      sequence: int64
    - name: ui_neg_objs_bbox
      sequence:
        sequence: float64
    - name: ui_neg_objs_str
      sequence: string
    - name: ui_pos_obj_screen_bbox
      sequence: float64
    - name: ui_pos_obj_view_bbox
      sequence: int64
    - name: ui_pos_obj_type
      dtype: int64
    - name: ui_pos_obj_input_str
      dtype: string
    - name: ui_pos_obj_desc_str
      dtype: string
    - name: verb_str
      dtype: string
    - name: action
      dtype: string
    - name: instr
      dtype: string
    - name: trace_id
      dtype: string
    - name: goal
      dtype: string
    - name: app
      dtype: string
    - name: screen_w
      dtype: int64
    - name: screen_h
      dtype: int64
    - name: vh_w
      dtype: int64
    - name: vh_h
      dtype: int64
    - name: scale_x
      dtype: float64
    - name: scale_y
      dtype: float64
  splits:
    - name: train_au_tu
      num_bytes: 10639722044.4
      num_examples: 10900
    - name: test_au_tu
      num_bytes: 1345252172.816
      num_examples: 1344
  download_size: 11297393454
  dataset_size: 11984974217.216
license: cc-by-4.0
task_categories:
  - visual-question-answering
language:
  - en
Dataset Card for "MoTIF-automation"

Please refer to paper A Dataset for Interactive Vision Language Navigation with Unknown Command Feasibility