README.md

---
dataset_info:
  features:
  - name: dataset_source
    dtype: string
  - name: example_type
    dtype: string
  - name: text_caption
    dtype: string
  - name: text_caption_list
    sequence: string
  - name: rewrite_text_caption
    dtype: string
  - name: rewrite_text_caption_list
    sequence: string
  - name: rewrite_caption_levels
    sequence:
    - name: caption
      dtype: string
    - name: segments
      sequence:
      - name: tag
        dtype: int32
      - name: text
        dtype: string
  - name: general_scene_caption
    dtype: string
  - name: is_condition_pose_real_signal
    dtype: bool
  - name: is_pose_to_generate_real_signal
    dtype: bool
  - name: is_real_image_real_signal
    dtype: bool
  - name: ori_condition_pose_image
    dtype: image
  - name: ori_pose_to_generate_image
    dtype: image
  - name: ori_real_image
    dtype: image
  - name: ID
    dtype: string
  - name: metadata
    dtype: string
  - name: person_idx
    dtype: int32
  - name: person_bounding_box
    dtype: string
  - name: person_bounding_box_in_order
    dtype: string
  - name: num_detected_people
    dtype: int32
  - name: notes
    dtype: string
  - name: interaction_label
    dtype: string
  - name: number_of_interacting_people
    dtype: int32
  splits:
  - name: train
    num_bytes: 216607878988.595
    num_examples: 104595
  download_size: 210262961216
  dataset_size: 216607878988.595
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---