metadata
dataset_info:
features:
- name: image
dtype: image
- name: id
dtype: string
- name: filename
dtype: string
- name: width
dtype: int64
- name: height
dtype: int64
- name: polygons
list:
- name: class_id
dtype: int64
- name: class_name
dtype: string
- name: points
sequence:
sequence: float64
- name: classes
sequence: int64
- name: class_names
sequence: string
splits:
- name: train
num_bytes: 54999997026
num_examples: 140000
download_size: 55721768952
dataset_size: 54999997026
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
Dataset processed from DocLayout-YOLO
Dataclass mapping
CLASS_NAMES = {
0: 'QR code', 1: 'advertisement', 2: 'algorithm', 3: 'answer', 4: 'author',
5: 'barcode', 6: 'bill', 7: 'blank', 8: 'bracket', 9: 'breakout',
10: 'byline', 11: 'caption', 12: 'catalogue', 13: 'chapter title', 14: 'code',
15: 'correction', 16: 'credit', 17: 'dateline', 18: 'drop cap', 19: "editor's note",
20: 'endnote', 21: 'examinee information', 22: 'fifth-level title', 23: 'figure', 24: 'first-level question number',
25: 'first-level title', 26: 'flag', 27: 'folio', 28: 'footer', 29: 'footnote',
30: 'formula', 31: 'fourth-level section title', 32: 'fourth-level title', 33: 'header', 34: 'headline',
35: 'index', 36: 'inside', 37: 'institute', 38: 'jump line', 39: 'kicker',
40: 'lead', 41: 'marginal note', 42: 'matching', 43: 'mugshot', 44: 'option',
45: 'ordered list', 46: 'other question number', 47: 'page number', 48: 'paragraph', 49: 'part',
50: 'play', 51: 'poem', 52: 'reference', 53: 'sealing line', 54: 'second-level question number',
55: 'second-level title', 56: 'section', 57: 'section title', 58: 'sidebar', 59: 'sub section title',
60: 'subhead', 61: 'subsub section title', 62: 'supplementary note', 63: 'table', 64: 'table caption',
65: 'table note', 66: 'teasers', 67: 'third-level question number', 68: 'third-level title', 69: 'title',
70: 'translator', 71: 'underscore', 72: 'unordered list', 73: 'weather forecast'
}