voice-biomarkers commited on
Commit
6f174d4
1 Parent(s): e1fda3e

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +27 -19
README.md CHANGED
@@ -22,38 +22,46 @@ size_categories:
22
  - 10K<n<100K
23
  ---
24
 
25
- # Kazakh Speech Corpus (KSC)
26
- ### Identifier: SLR102
27
 
28
  Source: https://www.openslr.org/140/
29
 
30
- Summary: A crowdsourced open-source Kazakh speech corpus developed by ISSAI (330 hours)
31
 
32
  Category: Speech
33
 
34
- License: Attribution 4.0 International (CC BY 4.0)
35
-
36
 
37
 
38
  About this resource:
39
 
40
- A crowdsourced open-source speech corpus for the Kazakh language. The KSC contains around 332 hours of transcribed audio comprising over 153,000 utterances spoken by participants from different regions and age groups, as well as both genders. It was carefully inspected by native Kazakh speakers to ensure high quality.
 
 
 
 
41
 
42
- The dataset is primarily intended to be used for training automatic speech recognition systems.
43
 
 
 
 
 
 
 
 
44
 
45
  To cite the dataset, please use the following BibTeX entry:
46
  ```
47
- @inproceedings{khassanov-etal-2021-crowdsourced,
48
- title = "A Crowdsourced Open-Source {K}azakh Speech Corpus and Initial Speech Recognition Baseline",
49
- author={Yerbolat Khassanov and Saida Mussakhojayeva and Almas Mirzakhmetov and Alen Adiyev and Mukhamet Nurpeiissov and Huseyin Atakan Varol},
50
- booktitle = "Proceedings of the 16th Conference of the European Chapter of the Association for Computational Linguistics: Main Volume",
51
- month = apr,
52
- year = "2021",
53
- address = "Online",
54
- publisher = "Association for Computational Linguistics",
55
- url = "https://aclanthology.org/2021.eacl-main.58",
56
- doi = "10.18653/v1/2021.eacl-main.58",
57
- pages = "697--706"
58
  }
59
- ```
 
22
  - 10K<n<100K
23
  ---
24
 
25
+ # Kazakh Speech Dataset (KSD)
26
+ ### Identifier: SLR140
27
 
28
  Source: https://www.openslr.org/140/
29
 
30
+ Summary: High-quality open source Kazakh speech corpus developed by the Department of Artificial Intelligence and Big Data of Al-Farabi Kazakh National University (554 hours)
31
 
32
  Category: Speech
33
 
34
+ License: Attribution-ShareAlike 3.0 Unported (CC BY-SA 3.0 US)
 
35
 
36
 
37
  About this resource:
38
 
39
+ High-quality open source Kazakh speech corpus.
40
+
41
+ The corpus contains about 554 hours of transcribed audio recordings, including 204250 utterances uttered by participants from different regions and age groups, as well as by both sexes.
42
+
43
+ All audio files were recorded using mobile devices (iOS and Android).
44
 
45
+ The corpus was selectively checked by native speakers of the Kazakh language to ensure high quality.
46
 
47
+ The data set is primarily intended for use in training systems for automatic speech recognition.
48
+
49
+ Technical characteristics of audio files: .wav format, 16 kB, 22 and 44 kHz.
50
+
51
+ The founders of the corpus:
52
+ - Nurgali Kadyrbek(https://orcid.org/0000-0002-5461-8899)
53
+ - Madina Mansurova(https://orcid.org/0000-0002-9680-2758)
54
 
55
  To cite the dataset, please use the following BibTeX entry:
56
  ```
57
+ @inproceedings{
58
+ mansurova-kadyrbek-2023-kazakh-speech-dataset,
59
+ title = "The Development of a Kazakh Speech Recognition Model Using a Convolutional Neural Network with Fixed Character Level Filters",
60
+ author = "Madina Mansurova and Nurgali Kadyrbek",
61
+ booktitle = "Proceedings of the Big Data and Cognitive Computing",
62
+ month = "July 20",
63
+ year = "2023",
64
+ pages = "5--9",
65
+ url = "https://doi.org/10.3390/bdcc7030132"
 
 
66
  }
67
+ ```