---
dataset_info:
  features:
  - name: image
    dtype: image
  - name: annotation
    dtype: image
  - name: semantic_class_to_id
    struct:
    - name: background
      dtype: int64
    - name: book
      dtype: int64
    - name: car
      dtype: int64
    - name: person
      dtype: int64
  splits:
  - name: train
    num_bytes: 74144581.0
    num_examples: 1600
  - name: validation
    num_bytes: 15357771.0
    num_examples: 330
  download_size: 88571489
  dataset_size: 89502352.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: validation
    path: data/validation-*
---