clovaai / cord

CORD: A Consolidated Receipt Dataset for Post-OCR Parsing

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Shield: CC BY 4.0

This work is licensed under a Creative Commons Attribution 4.0 International License.

CC BY 4.0

CORD: A Consolidated Receipt Dataset for Post-OCR Parsing

We introduce a novel dataset called CORD, which stands for a COnsolidated Receipt Dataset for post-OCR parsing.

teaser

Abstract [paper]

OCR is inevitably linked to NLP since its final output is in text. Advances in document intelligence are driving the need for a unified technology that integrates OCR with various NLP tasks, especially semantic parsing. Since OCR and semantic parsing have been studied as separate tasks so far, the datasets for each task on their own are rich, while those for the integrated post-OCR parsing tasks are relatively insufficient. In this study, we publish a consolidated dataset for receipt parsing as the first step towards post-OCR parsing tasks. The dataset consists of thousands of Indonesian receipts, which contains images and box/text annotations for OCR, and multi-level semantic labels for parsing. The proposed dataset can be used to address various OCR and parsing tasks.

Updates

  • CORD v2 data has been uploaded to the Hugging Face Datasets. We investigated all data and corrected the incorrect labels. Also, we added the attribute sub_group_id to each element of valid_line. We can use this information to describe more accurate hierarchy of the resulting parse. See the gt_parse of the examples containing the menu.sub_nm, and compare those of the CORD v1. [20220720]
  • CORD v1 data has been uploaded to the Hugging Face Datasets. CORD v1 has the same contents as v0 except the gt_parse attribute. gt_parse represents the parse format constructed from the valid_line. [20220720]
  • 1,000 sample dataset will be available soon. Some class labels shown in the original paper were removed due to Indonesian legal issues. In particular, the store_info, payment_info, and etc fields have been removed from the target class to be published. [20191212]
  • 1,000 sample dataset has been released. [20191226]
  • Some categories not used in the current dataset have been removed from the class definition. [20200210]

Key Features

  • Large Scale: over 11,000 Indonesian receipts collected from shops and restaurants
  • Fine-grained classes: five superclass and 42 subclass labels
  • Multi hierarchy: includes group annotations
  • Additional information: line group (row_id), region of interest (roi), cut lines (repeating_symbol), and is_key flag

Data Specification (for the whole dataset)

Class Definition (total 30)

No Category Tag field (subclasses) Description
1 menu (14) menu.nm name of menu
2 menu.num identification # of menu
3 menu.unitprice unit price of menu
4 menu.cnt quantity of menu
5 menu.discountprice discounted price of menu
6 menu.price total price of menu
7 menu.itemsubtotal price of each menu after discount applied
8 menu.vatyn whether the price includes tax or not
9 menu.etc others
10 menu.sub_nm name of submenu
11 menu.sub_num identification # of submenu
12 menu.sub_unitprice unit price of submenu
13 menu.sub_cnt quantity of submenu
14 menu.sub_discountprice discounted price of submenu
15 menu.sub_price total price of submenu
16 menu.sub_etc others
17 void menu (2) void_menu.nm name of menu
18 voidmenu.num identification # of menu
19 voidmenu.unitprice unit price of menu
20 voidmenu.cnt quantity of menu
21 void_menu.price total price of menu
22 voidmenu.etc others
23 subtotal (6) subtotal.subtotal_price subtotal price
24 subtotal.discount_price discounted price in total
25 subtotal.subtotal_count Total number of items
26 subtotal.service_price service charge
27 subtotal.othersvc_price added charge other than service charge
28 subtotal.tax_price tax amount
29 subtotal.tax_and_service tax + service
30 subtotal.etc others
31 void total (0) voidtotal.subtotal_price void subtotal price
32 voidtotal.tax_price void tax price
33 voidtotal.total_price total void price
34 voidtotal.etc void etc information
35 total (8) total.total_price total price
36 total.total_etc others
37 total.cashprice amount of price paid in cash
38 total.changeprice amount of change in cash
39 total.creditcardprice amount of price paid in credit/debit card
40 total.emoneyprice amount of price paid in emoney, point
41 total.menutype_cnt total count of type of menu
42 total.menuqty_cnt total count of quantity

Json Hierarchy

Attribute Name Description
valid_line words quad Four coordinates of quadrilateral
is_key Flag to indicates the text used as a key or not
row_id Line index
text Incorporating text of the corresponding box
category Parse class label
group_id Group id to which the valid_line belongs
---------------- ---------- ------ ----------------------------------------------------------
meta version Dataset version
image_id Corresponding image id
split 'train' or 'dev' or 'test'
image_size Size of the image (by pixel)
---------------- ---------- ------ ----------------------------------------------------------
roi* Four coordinates that encompass the area of receipt region
---------------- ---------- ------ ----------------------------------------------------------
repeating_symbol quad Four coordinates of quadrilateral
text = or - or . or etc.

*A blank 'roi' value means the entire area of the image.

Download Link

Version Name Total # train # dev # test release date
v0 sample (zip) 1,000 800 100 100 26 Dec 2019
v1 Hugging Face Datasets Link 1,000 800 100 100 20 Jul 2022
v2 Hugging Face Datasets Link 1,000 800 100 100 20 Jul 2022

Citation

CORD: A Consolidated Receipt Dataset for Post-OCR Parsing

@article{park2019cord,
  title={CORD: A Consolidated Receipt Dataset for Post-OCR Parsing},
  author={Park, Seunghyun and Shin, Seung and Lee, Bado and Lee, Junyeop and Surh, Jaeheung and Seo, Minjoon and Lee, Hwalsuk}
  booktitle={Document Intelligence Workshop at Neural Information Processing Systems}
  year={2019}
}

Post-OCR parsing: building simple and robust parser via BIO tagging

@article{hwang2019post,
  title={Post-OCR parsing: building simple and robust parser via BIO tagging},
  author={Hwang, Wonseok and Kim, Seonghyeon and Yim, Jinyeong and Seo, Minjoon and Park, Seunghyun and Park, Sungrae and Lee, Junyeop and Lee, Bado and Lee, Hwalsuk}
  booktitle={Document Intelligence Workshop at Neural Information Processing Systems}
  year={2019}
}

OCR-free Document Understanding Transformer 🍩

@article{kim2021donut,
   title={OCR-free Document Understanding Transformer},
   author={Kim, Geewook and Hong, Teakgyu and Yim, Moonbin and Nam, JeongYeon and Park, Jinyoung and Yim, Jinyeong and Hwang, Wonseok and Yun, Sangdoo and Han, Dongyoon and Park, Seunghyun},
   journal={arXiv preprint arXiv:2111.15664},
   year={2021}
}

About

CORD: A Consolidated Receipt Dataset for Post-OCR Parsing

License:Creative Commons Attribution 4.0 International