jnaiman commited on
Commit
f87fbcc
·
1 Parent(s): aff40a1

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +3 -1
README.md CHANGED
@@ -13,10 +13,12 @@ size_categories:
13
 
14
  Over 1.5 Million synthetically generated ground-truth/[OCR](https://en.wikipedia.org/wiki/Optical_character_recognition) pairs for post correction tasks from our paper "[Large Synthetic Data from the ar𝜒iv for OCR Post Correction of Historic Scientific Articles](https://dl.acm.org/doi/10.1007/978-3-031-43849-3_23)".
15
 
16
- Synthetic ground truth sentences have been mined from the [ar𝜒iv Bulk Downloads](https://info.arxiv.org/help/bulk_data/index.html) source documents,
17
  and Optical Character Recognition (OCR)
18
  sentences have been generated with the [Tesseract](https://github.com/tesseract-ocr/tesseract) OCR engine on the PDF pages generated from compiled source documents.
19
 
 
 
20
  No page augmentation has been applied to any of the PDF documents (i.e. these are "clean" pages without warping, dust, etc.)
21
 
22
  ## Resources
 
13
 
14
  Over 1.5 Million synthetically generated ground-truth/[OCR](https://en.wikipedia.org/wiki/Optical_character_recognition) pairs for post correction tasks from our paper "[Large Synthetic Data from the ar𝜒iv for OCR Post Correction of Historic Scientific Articles](https://dl.acm.org/doi/10.1007/978-3-031-43849-3_23)".
15
 
16
+ Synthetic ground truth (SGT) sentences have been mined from the [ar𝜒iv Bulk Downloads](https://info.arxiv.org/help/bulk_data/index.html) source documents,
17
  and Optical Character Recognition (OCR)
18
  sentences have been generated with the [Tesseract](https://github.com/tesseract-ocr/tesseract) OCR engine on the PDF pages generated from compiled source documents.
19
 
20
+ SGT/OCR pairs come from astronomy articles in the years 1991-2011.
21
+
22
  No page augmentation has been applied to any of the PDF documents (i.e. these are "clean" pages without warping, dust, etc.)
23
 
24
  ## Resources