Datasets:
Update README.md
Browse files
README.md
CHANGED
@@ -18,4 +18,64 @@ configs:
|
|
18 |
data_files:
|
19 |
- split: train
|
20 |
path: data/train-*
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
21 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
18 |
data_files:
|
19 |
- split: train
|
20 |
path: data/train-*
|
21 |
+
license: eupl-1.1
|
22 |
+
task_categories:
|
23 |
+
- automatic-speech-recognition
|
24 |
+
- text-to-speech
|
25 |
+
- translation
|
26 |
+
language:
|
27 |
+
- ga
|
28 |
+
- en
|
29 |
+
size_categories:
|
30 |
+
- 10K<n<100K
|
31 |
---
|
32 |
+
|
33 |
+
# Dataset Details
|
34 |
+
|
35 |
+
Synthetic audio dataset, created using Azure text-to-speech service.
|
36 |
+
The bilingual text is a portion of the EUbookshop dataset, consisting of 33,634 text segments.
|
37 |
+
The dataset includes two sets of audio data, one with a female voice (OrlaNeural) and the other with a male voice (ColmNeural).
|
38 |
+
The speech data comprises approximately 159 hours and 45 minutes (159:45:05) spread across 67,268 utterances.
|
39 |
+
|
40 |
+
|
41 |
+
## Dataset Structure
|
42 |
+
|
43 |
+
```
|
44 |
+
Dataset({
|
45 |
+
features: ['audio', 'text_ga', 'text_en'],
|
46 |
+
num_rows: 67268
|
47 |
+
})
|
48 |
+
```
|
49 |
+
|
50 |
+
|
51 |
+
## Citations
|
52 |
+
|
53 |
+
* This speech dataset `EUbookshop-Speech-Irish` was created and introduced as part of the following paper:
|
54 |
+
|
55 |
+
```
|
56 |
+
@inproceedings{moslem2024leveraging,
|
57 |
+
title={Leveraging Synthetic Audio Data for End-to-End Low-Resource Speech Translation},
|
58 |
+
author={Moslem, Yasmin},
|
59 |
+
booktitle={Proceedings of the 2024 International Conference on Spoken Language Translation (IWSLT 2024)},
|
60 |
+
year={2024},
|
61 |
+
month={April}
|
62 |
+
}
|
63 |
+
```
|
64 |
+
|
65 |
+
* The original EUbookshop bilingual text dataset was collected by the OPUS project, introduced in the following paper:
|
66 |
+
|
67 |
+
```
|
68 |
+
@INPROCEEDINGS{Tiedemann2012-OPUS,
|
69 |
+
title = "{Parallel Data, Tools and Interfaces in {OPUS}}",
|
70 |
+
booktitle = "{Proceedings of the Eighth International Conference on Language
|
71 |
+
Resources and Evaluation ({LREC}'12)}",
|
72 |
+
author = "Tiedemann, J{\"o}rg",
|
73 |
+
publisher = "European Language Resources Association (ELRA)",
|
74 |
+
pages = "2214--2218",
|
75 |
+
month = may,
|
76 |
+
year = 2012,
|
77 |
+
url = "http://www.lrec-conf.org/proceedings/lrec2012/pdf/463_Paper.pdf",
|
78 |
+
address = "Istanbul, Turkey"
|
79 |
+
}
|
80 |
+
|
81 |
+
```
|