mboillet starride-teklia commited on
Commit
c28cd7a
1 Parent(s): c763036

Update readme (#4)

Browse files

- Update readme (57b1d36bc7e8d6a89b59dadf987cb4a9849d135d)


Co-authored-by: Solène Tarride <starride-teklia@users.noreply.huggingface.co>

Files changed (1) hide show
  1. README.md +26 -15
README.md CHANGED
@@ -4,22 +4,27 @@ language:
4
  - fr
5
  task_categories:
6
  - image-to-text
7
- pretty_name: PyLaia RIMES
8
-
9
  dataset_info:
10
  features:
11
- - name: image
12
- dtype: image
13
- - name: text
14
- dtype: string
15
  splits:
16
- - name: train
17
- num_examples: 10188
18
- - name: validation
19
- num_examples: 1138
20
- - name: test
21
- num_examples: 778
22
  dataset_size: 12104
 
 
 
 
 
 
23
  ---
24
 
25
  # PyLaia RIMES Dataset
@@ -35,7 +40,7 @@ dataset_info:
35
 
36
  ## Dataset Description
37
 
38
- - **Homepage:** [ARTEMIS](https://artemis.telecom-sudparis.eu/2012/10/05/rimes/)
39
  - **PapersWithCode:** [Papers using the RIMES dataset](https://paperswithcode.com/dataset/rimes)
40
  - **Point of Contact:** [TEKLIA](https://teklia.com)
41
 
@@ -43,6 +48,12 @@ dataset_info:
43
 
44
  The RIMES database (Recognition and Indexation of handwritten documents and faxes) was created to evaluate automatic recognition and indexing systems for handwritten letters.
45
 
 
 
 
 
 
 
46
  ### Languages
47
 
48
  All the documents in the dataset are written in French.
@@ -61,5 +72,5 @@ All the documents in the dataset are written in French.
61
  ### Data Fields
62
 
63
 
64
- - `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].
65
- - `text`: the label transcription of the image.
 
4
  - fr
5
  task_categories:
6
  - image-to-text
7
+ pretty_name: RIMES-2011-line
 
8
  dataset_info:
9
  features:
10
+ - name: image
11
+ dtype: image
12
+ - name: text
13
+ dtype: string
14
  splits:
15
+ - name: train
16
+ num_examples: 10188
17
+ - name: validation
18
+ num_examples: 1138
19
+ - name: test
20
+ num_examples: 778
21
  dataset_size: 12104
22
+ tags:
23
+ - atr
24
+ - htr
25
+ - ocr
26
+ - modern
27
+ - handwritten
28
  ---
29
 
30
  # PyLaia RIMES Dataset
 
40
 
41
  ## Dataset Description
42
 
43
+ - **Homepage:** [Zenodo](https://zenodo.org/records/10812725)
44
  - **PapersWithCode:** [Papers using the RIMES dataset](https://paperswithcode.com/dataset/rimes)
45
  - **Point of Contact:** [TEKLIA](https://teklia.com)
46
 
 
48
 
49
  The RIMES database (Recognition and Indexation of handwritten documents and faxes) was created to evaluate automatic recognition and indexing systems for handwritten letters.
50
 
51
+ The database was collected by asking volunteers to write handwritten letters in exchange for gift certificates. Volunteers were given a fictitious identity (same gender as the real one) and up to 5 scenarios. Each scenario was chosen from among 9 realistic topics: change of personal data (address, bank account), request for information, opening and closing (customer account), change of contract or order, complaint (poor quality of service...), payment difficulties (request for delay, tax exemption...), reminder, complaint with other circumstances and a target (administrations or service providers (telephone, electricity, bank, insurance). The volunteers wrote a letter with this information in their own words. The layout was free and the only request was to use white paper and write legibly in black ink.
52
+
53
+ The campaign was a success, with more than 1,300 people contributing to the RIMES database by writing up to 5 letters. The resulting RIMES database contains 12,723 pages, corresponding to 5605 mails of two to three pages each.
54
+
55
+ Note that all images are resized to a fixed height of 128 pixels.
56
+
57
  ### Languages
58
 
59
  All the documents in the dataset are written in French.
 
72
  ### Data Fields
73
 
74
 
75
+ - `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].
76
+ - `text`: the label transcription of the image.