mboillet commited on
Commit
f42209a
1 Parent(s): da24291

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +11 -3
README.md CHANGED
@@ -17,12 +17,18 @@ dataset_info:
17
  - name: validation
18
  num_examples: 4379
19
  dataset_size: 55967
 
 
 
 
 
 
20
  ---
21
 
22
- # NewsEye Dataset
23
 
24
  ## Table of Contents
25
- - [NewsEye Dataset](#newseye-dataset)
26
  - [Table of Contents](#table-of-contents)
27
  - [Dataset Description](#dataset-description)
28
  - [Languages](#languages)
@@ -44,6 +50,8 @@ The dataset comprises Austrian newspaper pages from 19th and early 20th century.
44
 
45
  The documents are in Austrian German with the Fraktur font.
46
 
 
 
47
  ## Dataset Structure
48
 
49
  ### Data Instances
@@ -58,5 +66,5 @@ The documents are in Austrian German with the Fraktur font.
58
  ### Data Fields
59
 
60
 
61
- - `image`: A PIL.Image.Image object containing the image. Note that when accessing the image column: dataset[0]["image"] the image file is automatically decoded. Decoding of a large number of image files might take a significant amount of time. Thus it is important to first query the sample index before the "image" column, i.e. dataset[0]["image"] should always be preferred over dataset["image"][0].
62
  - `text`: the label transcription of the image.
 
17
  - name: validation
18
  num_examples: 4379
19
  dataset_size: 55967
20
+ tags:
21
+ - atr
22
+ - htr
23
+ - ocr
24
+ - historical
25
+ - printed
26
  ---
27
 
28
+ # NewsEye Austrian - line level
29
 
30
  ## Table of Contents
31
+ - [NewsEye Austrian - line level](#newseye-austrian-line-level)
32
  - [Table of Contents](#table-of-contents)
33
  - [Dataset Description](#dataset-description)
34
  - [Languages](#languages)
 
50
 
51
  The documents are in Austrian German with the Fraktur font.
52
 
53
+ Note that all images are resized to a fixed height of 128 pixels.
54
+
55
  ## Dataset Structure
56
 
57
  ### Data Instances
 
66
  ### Data Fields
67
 
68
 
69
+ - `image`: a PIL.Image.Image object containing the image. Note that when accessing the image column (using dataset[0]["image"]), the image file is automatically decoded. Decoding of a large number of image files might take a significant amount of time. Thus it is important to first query the sample index before the "image" column, i.e. dataset[0]["image"] should always be preferred over dataset["image"][0].
70
  - `text`: the label transcription of the image.