publaynet.yaml

# Copyright 2021 The MLX Contributors
#
# SPDX-License-Identifier: Apache-2.0
id: publaynet
name: PubLayNet
description: PubLayNet is a large dataset of document images from PubMed Central Open Access Subset. Each document’s layout is annotated with both bounding boxes and polygonal segmentations.
version: 1.0.0
created: 2019-08-07
updated: 2019-08-07
format:
  - type: JPG
    url: https://en.wikipedia.org/wiki/JPEG
  - type: JSON
    url: https://json.org/
domain: Computer Vision

# Information about the entity that makes the data set available
provider:
  name: Data Asset eXchange
  url: https://developer.ibm.com/exchanges/data/all/publaynet/

# identifies where the data set is stored and how it is stored (REQUIRED)
repository:
  type: HTTP
  url: https://dax-cdn.cdn.appdomain.cloud/dax-publaynet/1.0.0/publaynet.tar.gz
  mime_type: application/x-tar
  sha_512: 087a58aeed533f953c041dc03e5f14bc0b1bf53c44fa1f5816fde7c0376174b1b67c837d93094f96fd6313aba172c914c7bc2d14c6f84b58913c0ddd550264eb
  size: 96G

# REQUIRED; data set license information
license:
  commercial: false
  name: CDLA-Permissive
  url: https://cdla.io/permissive-1-0/

# REQUIRED; describes relevant files in the data set archive
content:
  - pattern: train/*
    description: Images in the training subset
    type: regex
  - pattern: val/*
    description: Images in the validation subset
    type: regex
  - pattern: test/*
    description: Images in the testing subset
    type: regex
  - pattern: test.json
    description: Annotations for training images
    type: file
  - pattern: val.json
    description: Annotations for validation images
    type: file

# OPTIONAL; Identifies where the data set was obtained from
source:
  name: Images of research papers from PubMed and annotations from IBM Research Australia.
  url: https://www.ncbi.nlm.nih.gov/pmc/tools/openftlist/
  authors:
    - name: Xu Zhong
      url: https://researcher.watson.ibm.com/researcher/view.php?person=au1-peter.zhong
    - name: Jianbin Tang
      url: https://researcher.watson.ibm.com/researcher/view.php?person=ibm-Elaheh.Shafieibavani
    - name: Antonio Jimeno Yepes
      url: https://researcher.watson.ibm.com/researcher/view.php?person=au1-antonio.jimeno

# OPTIONAL; but recommended
seo_tags:
  - Document Layout Analysis
  - Text

# OPTIONAL; assets that complement this data set, e.g. notebooks
related_assets:
  - name: Publaynet Watson Studio project
    description: Watson Studio Gallery project for the publaynet data set
    mime_type: text/html
    url: https://dataplatform.cloud.ibm.com/analytics/notebooks/v2/43cb95d9-6c3e-479c-a189-8c9ff3524ec1/view?access_token=bb8ce645cf114b5f5512ae2eb9c7badcf0927f313e8f76b8138d0701289484e6
  - name: "Image-based table recognition: data, model, and evaluation"
    description: Research paper
    mime_type: text/html
    url: https://arxiv.org/abs/1911.10683

# OPTIONAL; url for the markdown file which describes the asset
readme_url: https://raw.githubusercontent.com/machine-learning-exchange/katalog/main/dataset-samples/publaynet/publaynet.md