Darayut commited on
Commit
e4c0104
·
verified ·
1 Parent(s): 414548b

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +8 -17
README.md CHANGED
@@ -2,6 +2,7 @@
2
  license: cc-by-4.0
3
  language:
4
  - km
 
5
  task_categories:
6
  - object-detection
7
  tags:
@@ -17,9 +18,9 @@ size_categories:
17
  - 10K<n<100K
18
  ---
19
 
20
- # Khmer Text Detection — Ultimate Dataset
21
 
22
- A large-scale, multi-source dataset for **Khmer text detection** using YOLO-format bounding box annotations.
23
  Combines real scene text, document layout images, and synthetically generated Khmer document images.
24
 
25
  ---
@@ -43,15 +44,15 @@ Combines real scene text, document layout images, and synthetically generated Kh
43
 
44
  ## Data Sources
45
 
46
- ### 1. Real Scene Text (`real_scene_text`)
47
  Images captured in natural environments — street signs, storefronts, billboards, and handwritten Khmer documents.
48
  Derived from the **DonkeySmall** base dataset (~26K images).
49
 
50
- ### 2. Document Layout (`doclaynet_khmer`)
51
- Images from the **DocLayNet**-style multi-class document layout corpus, re-labelled for the two-class schema.
52
- Covers printed Khmer documents: official reports, newspapers, and books.
53
 
54
- ### 3. Synthetic Khmer Documents (`synthetic_khmer_doc`)
55
  Programmatically generated document images using custom Khmer text rendering pipelines.
56
  Fonts, sizes, backgrounds, and layouts are randomised. Labels are auto-generated (zero annotation cost).
57
 
@@ -146,13 +147,3 @@ names:
146
  [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/) — free to use, share, and adapt with attribution.
147
 
148
  ---
149
-
150
- ## Citation
151
-
152
- ```bibtex
153
- @dataset{khmer_text_detection_ultimate,
154
- title = {Khmer Text Detection — Ultimate Dataset},
155
- year = {2025},
156
- url = {https://huggingface.co/datasets/Darayut/Multilingual-Textline-Detection-Dataset}
157
- }
158
- ```
 
2
  license: cc-by-4.0
3
  language:
4
  - km
5
+ - en
6
  task_categories:
7
  - object-detection
8
  tags:
 
18
  - 10K<n<100K
19
  ---
20
 
21
+ # Textline Detection — Ultimate Dataset
22
 
23
+ A large-scale, multi-source dataset for **textline detection** using YOLO-format bounding box annotations.
24
  Combines real scene text, document layout images, and synthetically generated Khmer document images.
25
 
26
  ---
 
44
 
45
  ## Data Sources
46
 
47
+ ### 1. Real Scene Text
48
  Images captured in natural environments — street signs, storefronts, billboards, and handwritten Khmer documents.
49
  Derived from the **DonkeySmall** base dataset (~26K images).
50
 
51
+ ### 2. Document Layout
52
+ Images from the **DocLayNet** multi-class document layout corpus, re-labelled for the two-class schema.
53
+ Covers documents: official reports, newspapers, and books.
54
 
55
+ ### 3. Synthetic Khmer Documents
56
  Programmatically generated document images using custom Khmer text rendering pipelines.
57
  Fonts, sizes, backgrounds, and layouts are randomised. Labels are auto-generated (zero annotation cost).
58
 
 
147
  [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/) — free to use, share, and adapt with attribution.
148
 
149
  ---