Datasets:
metadata
license: cc-by-4.0
task_categories:
- object-detection
language:
- as
- bn
- en
- gu
- hi
- kn
- ml
- mr
- or
- pa
- ta
- te
- ur
tags:
- document-layout
- indic
- ocr
- layout-analysis
dataset_info:
features:
- name: image
dtype: image
- name: file_name
dtype: string
- name: language
dtype: string
- name: language_code
dtype: string
- name: domain
dtype: string
- name: domain_code
dtype: string
- name: width
dtype: int32
- name: height
dtype: int32
- name: objects
struct:
- name: label
sequence: string
- name: label_id
sequence: int64
- name: bbox
sequence:
sequence: float32
configs:
- config_name: default
data_files:
- split: train
path: data/*/train-*.parquet
- split: validation
path: data/*/validation-*.parquet
- split: test
path: data/*/test-*.parquet
features:
- name: image
dtype: image
- name: file_name
dtype: string
- name: language
dtype: string
- name: language_code
dtype: string
- name: domain
dtype: string
- name: domain_code
dtype: string
- name: width
dtype: int32
- name: height
dtype: int32
- name: objects
struct:
- name: label
sequence: string
- name: label_id
sequence: int64
- name: bbox
sequence:
sequence: float32
- config_name: assamese
data_files:
- split: train
path: data/as-*/train-*.parquet
- split: validation
path: data/as-*/validation-*.parquet
- split: test
path: data/as-*/test-*.parquet
features:
- name: image
dtype: image
- name: file_name
dtype: string
- name: language
dtype: string
- name: language_code
dtype: string
- name: domain
dtype: string
- name: domain_code
dtype: string
- name: width
dtype: int32
- name: height
dtype: int32
- name: objects
struct:
- name: label
sequence: string
- name: label_id
sequence: int64
- name: bbox
sequence:
sequence: float32
- config_name: bengali
data_files:
- split: train
path: data/bn-*/train-*.parquet
- split: validation
path: data/bn-*/validation-*.parquet
- split: test
path: data/bn-*/test-*.parquet
features:
- name: image
dtype: image
- name: file_name
dtype: string
- name: language
dtype: string
- name: language_code
dtype: string
- name: domain
dtype: string
- name: domain_code
dtype: string
- name: width
dtype: int32
- name: height
dtype: int32
- name: objects
struct:
- name: label
sequence: string
- name: label_id
sequence: int64
- name: bbox
sequence:
sequence: float32
- config_name: bodo
data_files:
- split: train
path: data/bd-*/train-*.parquet
- split: validation
path: data/bd-*/validation-*.parquet
- split: test
path: data/bd-*/test-*.parquet
features:
- name: image
dtype: image
- name: file_name
dtype: string
- name: language
dtype: string
- name: language_code
dtype: string
- name: domain
dtype: string
- name: domain_code
dtype: string
- name: width
dtype: int32
- name: height
dtype: int32
- name: objects
struct:
- name: label
sequence: string
- name: label_id
sequence: int64
- name: bbox
sequence:
sequence: float32
- config_name: dogri
data_files:
- split: train
path: data/do-*/train-*.parquet
- split: validation
path: data/do-*/validation-*.parquet
- split: test
path: data/do-*/test-*.parquet
features:
- name: image
dtype: image
- name: file_name
dtype: string
- name: language
dtype: string
- name: language_code
dtype: string
- name: domain
dtype: string
- name: domain_code
dtype: string
- name: width
dtype: int32
- name: height
dtype: int32
- name: objects
struct:
- name: label
sequence: string
- name: label_id
sequence: int64
- name: bbox
sequence:
sequence: float32
- config_name: english
data_files:
- split: train
path: data/en-*/train-*.parquet
- split: validation
path: data/en-*/validation-*.parquet
- split: test
path: data/en-*/test-*.parquet
features:
- name: image
dtype: image
- name: file_name
dtype: string
- name: language
dtype: string
- name: language_code
dtype: string
- name: domain
dtype: string
- name: domain_code
dtype: string
- name: width
dtype: int32
- name: height
dtype: int32
- name: objects
struct:
- name: label
sequence: string
- name: label_id
sequence: int64
- name: bbox
sequence:
sequence: float32
- config_name: gujarati
data_files:
- split: train
path: data/gu-*/train-*.parquet
- split: validation
path: data/gu-*/validation-*.parquet
- split: test
path: data/gu-*/test-*.parquet
features:
- name: image
dtype: image
- name: file_name
dtype: string
- name: language
dtype: string
- name: language_code
dtype: string
- name: domain
dtype: string
- name: domain_code
dtype: string
- name: width
dtype: int32
- name: height
dtype: int32
- name: objects
struct:
- name: label
sequence: string
- name: label_id
sequence: int64
- name: bbox
sequence:
sequence: float32
- config_name: hindi
data_files:
- split: train
path: data/hi-*/train-*.parquet
- split: validation
path: data/hi-*/validation-*.parquet
- split: test
path: data/hi-*/test-*.parquet
features:
- name: image
dtype: image
- name: file_name
dtype: string
- name: language
dtype: string
- name: language_code
dtype: string
- name: domain
dtype: string
- name: domain_code
dtype: string
- name: width
dtype: int32
- name: height
dtype: int32
- name: objects
struct:
- name: label
sequence: string
- name: label_id
sequence: int64
- name: bbox
sequence:
sequence: float32
- config_name: kannada
data_files:
- split: train
path: data/kn-*/train-*.parquet
- split: validation
path: data/kn-*/validation-*.parquet
- split: test
path: data/kn-*/test-*.parquet
features:
- name: image
dtype: image
- name: file_name
dtype: string
- name: language
dtype: string
- name: language_code
dtype: string
- name: domain
dtype: string
- name: domain_code
dtype: string
- name: width
dtype: int32
- name: height
dtype: int32
- name: objects
struct:
- name: label
sequence: string
- name: label_id
sequence: int64
- name: bbox
sequence:
sequence: float32
- config_name: kashmiri
data_files:
- split: train
path: data/ks-*/train-*.parquet
- split: validation
path: data/ks-*/validation-*.parquet
- split: test
path: data/ks-*/test-*.parquet
features:
- name: image
dtype: image
- name: file_name
dtype: string
- name: language
dtype: string
- name: language_code
dtype: string
- name: domain
dtype: string
- name: domain_code
dtype: string
- name: width
dtype: int32
- name: height
dtype: int32
- name: objects
struct:
- name: label
sequence: string
- name: label_id
sequence: int64
- name: bbox
sequence:
sequence: float32
- config_name: konkani
data_files:
- split: train
path: data/kk-*/train-*.parquet
- split: validation
path: data/kk-*/validation-*.parquet
- split: test
path: data/kk-*/test-*.parquet
features:
- name: image
dtype: image
- name: file_name
dtype: string
- name: language
dtype: string
- name: language_code
dtype: string
- name: domain
dtype: string
- name: domain_code
dtype: string
- name: width
dtype: int32
- name: height
dtype: int32
- name: objects
struct:
- name: label
sequence: string
- name: label_id
sequence: int64
- name: bbox
sequence:
sequence: float32
- config_name: malayalam
data_files:
- split: train
path: data/ml-*/train-*.parquet
- split: validation
path: data/ml-*/validation-*.parquet
- split: test
path: data/ml-*/test-*.parquet
features:
- name: image
dtype: image
- name: file_name
dtype: string
- name: language
dtype: string
- name: language_code
dtype: string
- name: domain
dtype: string
- name: domain_code
dtype: string
- name: width
dtype: int32
- name: height
dtype: int32
- name: objects
struct:
- name: label
sequence: string
- name: label_id
sequence: int64
- name: bbox
sequence:
sequence: float32
- config_name: marathi
data_files:
- split: train
path: data/mr-*/train-*.parquet
- split: validation
path: data/mr-*/validation-*.parquet
- split: test
path: data/mr-*/test-*.parquet
features:
- name: image
dtype: image
- name: file_name
dtype: string
- name: language
dtype: string
- name: language_code
dtype: string
- name: domain
dtype: string
- name: domain_code
dtype: string
- name: width
dtype: int32
- name: height
dtype: int32
- name: objects
struct:
- name: label
sequence: string
- name: label_id
sequence: int64
- name: bbox
sequence:
sequence: float32
- config_name: maithili
data_files:
- split: train
path: data/mt-*/train-*.parquet
- split: validation
path: data/mt-*/validation-*.parquet
- split: test
path: data/mt-*/test-*.parquet
features:
- name: image
dtype: image
- name: file_name
dtype: string
- name: language
dtype: string
- name: language_code
dtype: string
- name: domain
dtype: string
- name: domain_code
dtype: string
- name: width
dtype: int32
- name: height
dtype: int32
- name: objects
struct:
- name: label
sequence: string
- name: label_id
sequence: int64
- name: bbox
sequence:
sequence: float32
- config_name: manipuri
data_files:
- split: train
path: data/mp-*/train-*.parquet
- split: validation
path: data/mp-*/validation-*.parquet
- split: test
path: data/mp-*/test-*.parquet
features:
- name: image
dtype: image
- name: file_name
dtype: string
- name: language
dtype: string
- name: language_code
dtype: string
- name: domain
dtype: string
- name: domain_code
dtype: string
- name: width
dtype: int32
- name: height
dtype: int32
- name: objects
struct:
- name: label
sequence: string
- name: label_id
sequence: int64
- name: bbox
sequence:
sequence: float32
- config_name: nepali
data_files:
- split: train
path: data/ne-*/train-*.parquet
- split: validation
path: data/ne-*/validation-*.parquet
- split: test
path: data/ne-*/test-*.parquet
features:
- name: image
dtype: image
- name: file_name
dtype: string
- name: language
dtype: string
- name: language_code
dtype: string
- name: domain
dtype: string
- name: domain_code
dtype: string
- name: width
dtype: int32
- name: height
dtype: int32
- name: objects
struct:
- name: label
sequence: string
- name: label_id
sequence: int64
- name: bbox
sequence:
sequence: float32
- config_name: odia
data_files:
- split: train
path: data/or-*/train-*.parquet
- split: validation
path: data/or-*/validation-*.parquet
- split: test
path: data/or-*/test-*.parquet
features:
- name: image
dtype: image
- name: file_name
dtype: string
- name: language
dtype: string
- name: language_code
dtype: string
- name: domain
dtype: string
- name: domain_code
dtype: string
- name: width
dtype: int32
- name: height
dtype: int32
- name: objects
struct:
- name: label
sequence: string
- name: label_id
sequence: int64
- name: bbox
sequence:
sequence: float32
- config_name: punjabi
data_files:
- split: train
path: data/pa-*/train-*.parquet
- split: validation
path: data/pa-*/validation-*.parquet
- split: test
path: data/pa-*/test-*.parquet
features:
- name: image
dtype: image
- name: file_name
dtype: string
- name: language
dtype: string
- name: language_code
dtype: string
- name: domain
dtype: string
- name: domain_code
dtype: string
- name: width
dtype: int32
- name: height
dtype: int32
- name: objects
struct:
- name: label
sequence: string
- name: label_id
sequence: int64
- name: bbox
sequence:
sequence: float32
- config_name: sanskrit
data_files:
- split: train
path: data/sa-*/train-*.parquet
- split: validation
path: data/sa-*/validation-*.parquet
- split: test
path: data/sa-*/test-*.parquet
features:
- name: image
dtype: image
- name: file_name
dtype: string
- name: language
dtype: string
- name: language_code
dtype: string
- name: domain
dtype: string
- name: domain_code
dtype: string
- name: width
dtype: int32
- name: height
dtype: int32
- name: objects
struct:
- name: label
sequence: string
- name: label_id
sequence: int64
- name: bbox
sequence:
sequence: float32
- config_name: sindhi
data_files:
- split: train
path: data/sd-*/train-*.parquet
- split: validation
path: data/sd-*/validation-*.parquet
- split: test
path: data/sd-*/test-*.parquet
features:
- name: image
dtype: image
- name: file_name
dtype: string
- name: language
dtype: string
- name: language_code
dtype: string
- name: domain
dtype: string
- name: domain_code
dtype: string
- name: width
dtype: int32
- name: height
dtype: int32
- name: objects
struct:
- name: label
sequence: string
- name: label_id
sequence: int64
- name: bbox
sequence:
sequence: float32
- config_name: santali
data_files:
- split: train
path: data/sn-*/train-*.parquet
- split: validation
path: data/sn-*/validation-*.parquet
- split: test
path: data/sn-*/test-*.parquet
features:
- name: image
dtype: image
- name: file_name
dtype: string
- name: language
dtype: string
- name: language_code
dtype: string
- name: domain
dtype: string
- name: domain_code
dtype: string
- name: width
dtype: int32
- name: height
dtype: int32
- name: objects
struct:
- name: label
sequence: string
- name: label_id
sequence: int64
- name: bbox
sequence:
sequence: float32
- config_name: tamil
data_files:
- split: train
path: data/ta-*/train-*.parquet
- split: validation
path: data/ta-*/validation-*.parquet
- split: test
path: data/ta-*/test-*.parquet
features:
- name: image
dtype: image
- name: file_name
dtype: string
- name: language
dtype: string
- name: language_code
dtype: string
- name: domain
dtype: string
- name: domain_code
dtype: string
- name: width
dtype: int32
- name: height
dtype: int32
- name: objects
struct:
- name: label
sequence: string
- name: label_id
sequence: int64
- name: bbox
sequence:
sequence: float32
- config_name: telugu
data_files:
- split: train
path: data/te-*/train-*.parquet
- split: validation
path: data/te-*/validation-*.parquet
- split: test
path: data/te-*/test-*.parquet
features:
- name: image
dtype: image
- name: file_name
dtype: string
- name: language
dtype: string
- name: language_code
dtype: string
- name: domain
dtype: string
- name: domain_code
dtype: string
- name: width
dtype: int32
- name: height
dtype: int32
- name: objects
struct:
- name: label
sequence: string
- name: label_id
sequence: int64
- name: bbox
sequence:
sequence: float32
- config_name: urdu
data_files:
- split: train
path: data/ur-*/train-*.parquet
- split: validation
path: data/ur-*/validation-*.parquet
- split: test
path: data/ur-*/test-*.parquet
features:
- name: image
dtype: image
- name: file_name
dtype: string
- name: language
dtype: string
- name: language_code
dtype: string
- name: domain
dtype: string
- name: domain_code
dtype: string
- name: width
dtype: int32
- name: height
dtype: int32
- name: objects
struct:
- name: label
sequence: string
- name: label_id
sequence: int64
- name: bbox
sequence:
sequence: float32
- config_name: acts-and-rules
data_files:
- split: train
path: data/*-ar/train-*.parquet
- split: validation
path: data/*-ar/validation-*.parquet
- split: test
path: data/*-ar/test-*.parquet
features:
- name: image
dtype: image
- name: file_name
dtype: string
- name: language
dtype: string
- name: language_code
dtype: string
- name: domain
dtype: string
- name: domain_code
dtype: string
- name: width
dtype: int32
- name: height
dtype: int32
- name: objects
struct:
- name: label
sequence: string
- name: label_id
sequence: int64
- name: bbox
sequence:
sequence: float32
- config_name: brochure
data_files:
- split: train
path: data/*-br/train-*.parquet
- split: validation
path: data/*-br/validation-*.parquet
- split: test
path: data/*-br/test-*.parquet
features:
- name: image
dtype: image
- name: file_name
dtype: string
- name: language
dtype: string
- name: language_code
dtype: string
- name: domain
dtype: string
- name: domain_code
dtype: string
- name: width
dtype: int32
- name: height
dtype: int32
- name: objects
struct:
- name: label
sequence: string
- name: label_id
sequence: int64
- name: bbox
sequence:
sequence: float32
- config_name: form
data_files:
- split: train
path: data/*-fm/train-*.parquet
- split: validation
path: data/*-fm/validation-*.parquet
- split: test
path: data/*-fm/test-*.parquet
features:
- name: image
dtype: image
- name: file_name
dtype: string
- name: language
dtype: string
- name: language_code
dtype: string
- name: domain
dtype: string
- name: domain_code
dtype: string
- name: width
dtype: int32
- name: height
dtype: int32
- name: objects
struct:
- name: label
sequence: string
- name: label_id
sequence: int64
- name: bbox
sequence:
sequence: float32
- config_name: magazine
data_files:
- split: train
path: data/*-mg/train-*.parquet
- split: validation
path: data/*-mg/validation-*.parquet
- split: test
path: data/*-mg/test-*.parquet
features:
- name: image
dtype: image
- name: file_name
dtype: string
- name: language
dtype: string
- name: language_code
dtype: string
- name: domain
dtype: string
- name: domain_code
dtype: string
- name: width
dtype: int32
- name: height
dtype: int32
- name: objects
struct:
- name: label
sequence: string
- name: label_id
sequence: int64
- name: bbox
sequence:
sequence: float32
- config_name: manual
data_files:
- split: train
path: data/*-mn/train-*.parquet
- split: validation
path: data/*-mn/validation-*.parquet
- split: test
path: data/*-mn/test-*.parquet
features:
- name: image
dtype: image
- name: file_name
dtype: string
- name: language
dtype: string
- name: language_code
dtype: string
- name: domain
dtype: string
- name: domain_code
dtype: string
- name: width
dtype: int32
- name: height
dtype: int32
- name: objects
struct:
- name: label
sequence: string
- name: label_id
sequence: int64
- name: bbox
sequence:
sequence: float32
- config_name: newspaper
data_files:
- split: train
path: data/*-np/train-*.parquet
- split: validation
path: data/*-np/validation-*.parquet
- split: test
path: data/*-np/test-*.parquet
features:
- name: image
dtype: image
- name: file_name
dtype: string
- name: language
dtype: string
- name: language_code
dtype: string
- name: domain
dtype: string
- name: domain_code
dtype: string
- name: width
dtype: int32
- name: height
dtype: int32
- name: objects
struct:
- name: label
sequence: string
- name: label_id
sequence: int64
- name: bbox
sequence:
sequence: float32
- config_name: notice
data_files:
- split: train
path: data/*-nt/train-*.parquet
- split: validation
path: data/*-nt/validation-*.parquet
- split: test
path: data/*-nt/test-*.parquet
features:
- name: image
dtype: image
- name: file_name
dtype: string
- name: language
dtype: string
- name: language_code
dtype: string
- name: domain
dtype: string
- name: domain_code
dtype: string
- name: width
dtype: int32
- name: height
dtype: int32
- name: objects
struct:
- name: label
sequence: string
- name: label_id
sequence: int64
- name: bbox
sequence:
sequence: float32
- config_name: novel
data_files:
- split: train
path: data/*-nv/train-*.parquet
- split: validation
path: data/*-nv/validation-*.parquet
- split: test
path: data/*-nv/test-*.parquet
features:
- name: image
dtype: image
- name: file_name
dtype: string
- name: language
dtype: string
- name: language_code
dtype: string
- name: domain
dtype: string
- name: domain_code
dtype: string
- name: width
dtype: int32
- name: height
dtype: int32
- name: objects
struct:
- name: label
sequence: string
- name: label_id
sequence: int64
- name: bbox
sequence:
sequence: float32
- config_name: question-paper
data_files:
- split: train
path: data/*-qp/train-*.parquet
- split: validation
path: data/*-qp/validation-*.parquet
- split: test
path: data/*-qp/test-*.parquet
features:
- name: image
dtype: image
- name: file_name
dtype: string
- name: language
dtype: string
- name: language_code
dtype: string
- name: domain
dtype: string
- name: domain_code
dtype: string
- name: width
dtype: int32
- name: height
dtype: int32
- name: objects
struct:
- name: label
sequence: string
- name: label_id
sequence: int64
- name: bbox
sequence:
sequence: float32
- config_name: research-paper
data_files:
- split: train
path: data/*-rp/train-*.parquet
- split: validation
path: data/*-rp/validation-*.parquet
- split: test
path: data/*-rp/test-*.parquet
features:
- name: image
dtype: image
- name: file_name
dtype: string
- name: language
dtype: string
- name: language_code
dtype: string
- name: domain
dtype: string
- name: domain_code
dtype: string
- name: width
dtype: int32
- name: height
dtype: int32
- name: objects
struct:
- name: label
sequence: string
- name: label_id
sequence: int64
- name: bbox
sequence:
sequence: float32
- config_name: syllabus
data_files:
- split: train
path: data/*-sy/train-*.parquet
- split: validation
path: data/*-sy/validation-*.parquet
- split: test
path: data/*-sy/test-*.parquet
features:
- name: image
dtype: image
- name: file_name
dtype: string
- name: language
dtype: string
- name: language_code
dtype: string
- name: domain
dtype: string
- name: domain_code
dtype: string
- name: width
dtype: int32
- name: height
dtype: int32
- name: objects
struct:
- name: label
sequence: string
- name: label_id
sequence: int64
- name: bbox
sequence:
sequence: float32
- config_name: textbook
data_files:
- split: train
path: data/*-tb/train-*.parquet
- split: validation
path: data/*-tb/validation-*.parquet
- split: test
path: data/*-tb/test-*.parquet
features:
- name: image
dtype: image
- name: file_name
dtype: string
- name: language
dtype: string
- name: language_code
dtype: string
- name: domain
dtype: string
- name: domain_code
dtype: string
- name: width
dtype: int32
- name: height
dtype: int32
- name: objects
struct:
- name: label
sequence: string
- name: label_id
sequence: int64
- name: bbox
sequence:
sequence: float32
IndicDLP Sample
Sample subset of the IndicDLP (Indic Document Layout Parser) dataset.
- 23 languages: Assamese, Bengali, Bodo, Dogri, English, Gujarati, Hindi, Kannada, Kashmiri, Konkani, Malayalam, Maithili, Manipuri, Marathi, Nepali, Odia, Punjabi, Sanskrit, Sindhi, Santali, Tamil, Telugu, Urdu
- 12 document domains: Acts & Rules, Brochure, Form, Magazine, Manual, Newspaper, Notice, Novel, Question Paper, Research Paper, Syllabus, Textbook
- 42 layout classes (bounding box annotations)
- Splits: train / validation / test
Configs (subsets)
Use the config dropdown in the viewer, or specify when loading:
Languages: assamese (Assamese), bengali (Bengali), bodo (Bodo), dogri (Dogri), english (English), gujarati (Gujarati), hindi (Hindi), kannada (Kannada), kashmiri (Kashmiri), konkani (Konkani), malayalam (Malayalam), marathi (Marathi), maithili (Maithili), manipuri (Manipuri), nepali (Nepali), odia (Odia), punjabi (Punjabi), sanskrit (Sanskrit), sindhi (Sindhi), santali (Santali), tamil (Tamil), telugu (Telugu), urdu (Urdu)
Domains: acts-and-rules (Acts & Rules), brochure (Brochure), form (Form), magazine (Magazine), manual (Manual), newspaper (Newspaper), notice (Notice), novel (Novel), question-paper (Question Paper), research-paper (Research Paper), syllabus (Syllabus), textbook (Textbook)
Usage
from datasets import load_dataset
ds = load_dataset("onath/dlp-sample") # all data
ds = load_dataset("onath/dlp-sample", "hindi") # Hindi only
ds = load_dataset("onath/dlp-sample", "form") # Forms only
ds = load_dataset("onath/dlp-sample", "hindi", split="train") # Hindi train split
Annotation format
objects.bbox is in normalized YOLO format: [cx, cy, w, h] where values are in [0, 1].
Layout classes
advertisement, answer, author, chapter-title, contact-info, dateline,
figure, figure-caption, first-level-question, flag, folio, footer,
footnote, formula, header, headline, index, jumpline, options,
ordered-list, page-number, paragraph, placeholder-text, quote, reference,
second-level-question, section-title, sidebar, sub-headline, sub-ordered-list,
sub-section-title, subsub-headline, subsub-ordered-list, subsub-section-title,
subsub-unordered-list, sub-unordered-list, table, table-caption,
table-of-contents, third-level-question, unordered-list, website-link