Datasets:
Languages:
Greek
Size:
10K - 100K
Tags:
speech_corpus
greek_language
text-to-speech
neural_TTS
speech_synthesis
under_resourced_languages
License:
Update README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,337 @@
|
|
| 1 |
-
---
|
| 2 |
-
license: cc-by-4.0
|
| 3 |
-
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: cc-by-4.0
|
| 3 |
+
---
|
| 4 |
+
# STOMA: A Multi-Speaker Greek Speech Corpus
|
| 5 |
+
|
| 6 |
+
STOMA is a new multi-speaker Greek speech corpus designed to advance research in text-to-speech (TTS) synthesis and related speech technologies for under-resourced languages. The corpus comprises approximately 23 hours of professionally recorded read speech from six native speakers (three male and three female), captured under controlled studio conditions using a dual-booth setup to ensure acoustic consistency and high signal quality. The spoken material was selected from the Text Bank of the Center for the Greek Language, specifically from texts corresponding to the B2, C1, and C2 proficiency levels of the Certification of Attainment in Greek, ensuring linguistically rich and pedagogically balanced content. All recordings were standardized to 44.1 kHz, 16-bit mono PCM format and processed through a hybrid quality control pipeline combining automated normalization and manual verification. To assess dataset quality, we trained state-of-the-art neural TTS systems based on the FastSpeech2 acoustic model and the HiFi-GAN vocoder, achieving natural and intelligible synthesized speech. The resulting corpus provides a publicly accessible, high-quality resource that supports both linguistic research and the development of modern speech synthesis systems in Greek.
|
| 7 |
+
|
| 8 |
+
<div align="center">
|
| 9 |
+
|
| 10 |
+
<strong>Table 1: Text statistics for the full text collection (Greek Harvard, B2, C1, and C2), recorded in its entirety by the two primary speakers.</strong>
|
| 11 |
+
|
| 12 |
+
<table>
|
| 13 |
+
<thead>
|
| 14 |
+
<tr>
|
| 15 |
+
<th>Statistic</th>
|
| 16 |
+
<th>Gr. Harvard</th>
|
| 17 |
+
<th>B2</th>
|
| 18 |
+
<th>C1</th>
|
| 19 |
+
<th>C2</th>
|
| 20 |
+
<th>Total</th>
|
| 21 |
+
</tr>
|
| 22 |
+
</thead>
|
| 23 |
+
<tbody>
|
| 24 |
+
<tr>
|
| 25 |
+
<td>Total Sentences</td>
|
| 26 |
+
<td align="center">720</td>
|
| 27 |
+
<td align="center">1,087</td>
|
| 28 |
+
<td align="center">1,571</td>
|
| 29 |
+
<td align="center">1,296</td>
|
| 30 |
+
<td align="center">4,674</td>
|
| 31 |
+
</tr>
|
| 32 |
+
<tr>
|
| 33 |
+
<td>Total Words</td>
|
| 34 |
+
<td align="center">5,539</td>
|
| 35 |
+
<td align="center">15,102</td>
|
| 36 |
+
<td align="center">25,263</td>
|
| 37 |
+
<td align="center">22,722</td>
|
| 38 |
+
<td align="center">68,626</td>
|
| 39 |
+
</tr>
|
| 40 |
+
<tr>
|
| 41 |
+
<td>Total Characters</td>
|
| 42 |
+
<td align="center">32,301</td>
|
| 43 |
+
<td align="center">95,119</td>
|
| 44 |
+
<td align="center">164,700</td>
|
| 45 |
+
<td align="center">152,904</td>
|
| 46 |
+
<td align="center">445,024</td>
|
| 47 |
+
</tr>
|
| 48 |
+
<tr>
|
| 49 |
+
<td>Distinct Words</td>
|
| 50 |
+
<td align="center">3,343</td>
|
| 51 |
+
<td align="center">5,606</td>
|
| 52 |
+
<td align="center">8,965</td>
|
| 53 |
+
<td align="center">8,275</td>
|
| 54 |
+
<td align="center">20,253</td>
|
| 55 |
+
</tr>
|
| 56 |
+
<tr>
|
| 57 |
+
<td>Mean Words per Sentence</td>
|
| 58 |
+
<td align="center">7.69</td>
|
| 59 |
+
<td align="center">13.89</td>
|
| 60 |
+
<td align="center">16.08</td>
|
| 61 |
+
<td align="center">17.53</td>
|
| 62 |
+
<td align="center">14.68</td>
|
| 63 |
+
</tr>
|
| 64 |
+
<tr>
|
| 65 |
+
<td>Min Words per Sentence</td>
|
| 66 |
+
<td align="center">5</td>
|
| 67 |
+
<td align="center">1</td>
|
| 68 |
+
<td align="center">1</td>
|
| 69 |
+
<td align="center">1</td>
|
| 70 |
+
<td align="center">1</td>
|
| 71 |
+
</tr>
|
| 72 |
+
<tr>
|
| 73 |
+
<td>Max Words per Sentence</td>
|
| 74 |
+
<td align="center">9</td>
|
| 75 |
+
<td align="center">32</td>
|
| 76 |
+
<td align="center">41</td>
|
| 77 |
+
<td align="center">39</td>
|
| 78 |
+
<td align="center">41</td>
|
| 79 |
+
</tr>
|
| 80 |
+
</tbody>
|
| 81 |
+
</table>
|
| 82 |
+
|
| 83 |
+
</div>
|
| 84 |
+
|
| 85 |
+
<div align="center">
|
| 86 |
+
|
| 87 |
+
<strong>Table 2: Text statistics for the reduced text subset selected from the full collection, used for recordings by the secondary speakers.</strong>
|
| 88 |
+
|
| 89 |
+
<table>
|
| 90 |
+
<thead>
|
| 91 |
+
<tr>
|
| 92 |
+
<th>Statistic</th>
|
| 93 |
+
<th>Gr. Harvard</th>
|
| 94 |
+
<th>B2</th>
|
| 95 |
+
<th>C1</th>
|
| 96 |
+
<th>C2</th>
|
| 97 |
+
<th>Subset Total</th>
|
| 98 |
+
</tr>
|
| 99 |
+
</thead>
|
| 100 |
+
<tbody>
|
| 101 |
+
<tr>
|
| 102 |
+
<td>Total Sentences</td>
|
| 103 |
+
<td align="center">720</td>
|
| 104 |
+
<td align="center">190</td>
|
| 105 |
+
<td align="center">194</td>
|
| 106 |
+
<td align="center">197</td>
|
| 107 |
+
<td align="center">1,301</td>
|
| 108 |
+
</tr>
|
| 109 |
+
<tr>
|
| 110 |
+
<td>Total Words</td>
|
| 111 |
+
<td align="center">5,539</td>
|
| 112 |
+
<td align="center">2,913</td>
|
| 113 |
+
<td align="center">3,057</td>
|
| 114 |
+
<td align="center">3,265</td>
|
| 115 |
+
<td align="center">14,774</td>
|
| 116 |
+
</tr>
|
| 117 |
+
<tr>
|
| 118 |
+
<td>Total Characters</td>
|
| 119 |
+
<td align="center">32,301</td>
|
| 120 |
+
<td align="center">18,286</td>
|
| 121 |
+
<td align="center">20,190</td>
|
| 122 |
+
<td align="center">22,110</td>
|
| 123 |
+
<td align="center">92,887</td>
|
| 124 |
+
</tr>
|
| 125 |
+
<tr>
|
| 126 |
+
<td>Distinct Words</td>
|
| 127 |
+
<td align="center">3,343</td>
|
| 128 |
+
<td align="center">1,519</td>
|
| 129 |
+
<td align="center">1,604</td>
|
| 130 |
+
<td align="center">1,676</td>
|
| 131 |
+
<td align="center">6,899</td>
|
| 132 |
+
</tr>
|
| 133 |
+
<tr>
|
| 134 |
+
<td>Mean Words per Sentence</td>
|
| 135 |
+
<td align="center">7.69</td>
|
| 136 |
+
<td align="center">15.33</td>
|
| 137 |
+
<td align="center">15.76</td>
|
| 138 |
+
<td align="center">16.57</td>
|
| 139 |
+
<td align="center">11.36</td>
|
| 140 |
+
</tr>
|
| 141 |
+
<tr>
|
| 142 |
+
<td>Min Words per Sentence</td>
|
| 143 |
+
<td align="center">5</td>
|
| 144 |
+
<td align="center">4</td>
|
| 145 |
+
<td align="center">3</td>
|
| 146 |
+
<td align="center">4</td>
|
| 147 |
+
<td align="center">3</td>
|
| 148 |
+
</tr>
|
| 149 |
+
<tr>
|
| 150 |
+
<td>Max Words per Sentence</td>
|
| 151 |
+
<td align="center">9</td>
|
| 152 |
+
<td align="center">28</td>
|
| 153 |
+
<td align="center">28</td>
|
| 154 |
+
<td align="center">30</td>
|
| 155 |
+
<td align="center">30</td>
|
| 156 |
+
</tr>
|
| 157 |
+
</tbody>
|
| 158 |
+
</table>
|
| 159 |
+
|
| 160 |
+
</div>
|
| 161 |
+
|
| 162 |
+
<div align="center">
|
| 163 |
+
|
| 164 |
+
<strong>Table 3: Speech statistics of the STOMA corpus per speaker and overall.</strong>
|
| 165 |
+
|
| 166 |
+
<table>
|
| 167 |
+
<thead>
|
| 168 |
+
<tr>
|
| 169 |
+
<th>Speaker ID</th>
|
| 170 |
+
<th>Clips</th>
|
| 171 |
+
<th>Total Duration</th>
|
| 172 |
+
<th>Mean Duration (s)</th>
|
| 173 |
+
<th>Min–Max (s)</th>
|
| 174 |
+
</tr>
|
| 175 |
+
</thead>
|
| 176 |
+
<tbody>
|
| 177 |
+
<tr>
|
| 178 |
+
<td>F</td>
|
| 179 |
+
<td align="center">4,674</td>
|
| 180 |
+
<td align="center">7:46:21</td>
|
| 181 |
+
<td align="center">5.99</td>
|
| 182 |
+
<td align="center">0.41–16.60</td>
|
| 183 |
+
</tr>
|
| 184 |
+
<tr>
|
| 185 |
+
<td>M</td>
|
| 186 |
+
<td align="center">4,674</td>
|
| 187 |
+
<td align="center">8:53:37</td>
|
| 188 |
+
<td align="center">6.85</td>
|
| 189 |
+
<td align="center">0.48–22.31</td>
|
| 190 |
+
</tr>
|
| 191 |
+
<tr>
|
| 192 |
+
<td>F1</td>
|
| 193 |
+
<td align="center">1,301</td>
|
| 194 |
+
<td align="center">1:46:58</td>
|
| 195 |
+
<td align="center">4.93</td>
|
| 196 |
+
<td align="center">1.65–15.08</td>
|
| 197 |
+
</tr>
|
| 198 |
+
<tr>
|
| 199 |
+
<td>F2</td>
|
| 200 |
+
<td align="center">1,301</td>
|
| 201 |
+
<td align="center">1:27:38</td>
|
| 202 |
+
<td align="center">4.04</td>
|
| 203 |
+
<td align="center">1.46–11.13</td>
|
| 204 |
+
</tr>
|
| 205 |
+
<tr>
|
| 206 |
+
<td>M1</td>
|
| 207 |
+
<td align="center">1,301</td>
|
| 208 |
+
<td align="center">1:47:06</td>
|
| 209 |
+
<td align="center">4.94</td>
|
| 210 |
+
<td align="center">1.64–14.99</td>
|
| 211 |
+
</tr>
|
| 212 |
+
<tr>
|
| 213 |
+
<td>M2</td>
|
| 214 |
+
<td align="center">1,301</td>
|
| 215 |
+
<td align="center">1:26:43</td>
|
| 216 |
+
<td align="center">4.00</td>
|
| 217 |
+
<td align="center">1.10–10.39</td>
|
| 218 |
+
</tr>
|
| 219 |
+
<tr>
|
| 220 |
+
<td><strong>Total</strong></td>
|
| 221 |
+
<td align="center"><strong>14,552</strong></td>
|
| 222 |
+
<td align="center"><strong>23:08:26</strong></td>
|
| 223 |
+
<td align="center"><strong>5.72</strong></td>
|
| 224 |
+
<td align="center"><strong>0.41–22.31</strong></td>
|
| 225 |
+
</tr>
|
| 226 |
+
</tbody>
|
| 227 |
+
</table>
|
| 228 |
+
|
| 229 |
+
</div>
|
| 230 |
+
|
| 231 |
+
|
| 232 |
+
<div align="center">
|
| 233 |
+
|
| 234 |
+
<strong>Table 4: Demographic information of the STOMA corpus speakers.</strong>
|
| 235 |
+
|
| 236 |
+
<table>
|
| 237 |
+
<thead>
|
| 238 |
+
<tr>
|
| 239 |
+
<th>Name & Surname</th>
|
| 240 |
+
<th>Speaker ID</th>
|
| 241 |
+
<th>Age</th>
|
| 242 |
+
<th>Sex</th>
|
| 243 |
+
<th>Region Raised</th>
|
| 244 |
+
</tr>
|
| 245 |
+
</thead>
|
| 246 |
+
<tbody>
|
| 247 |
+
<tr>
|
| 248 |
+
<td align="center">Michail Raptakis</td>
|
| 249 |
+
<td align="center">M</td>
|
| 250 |
+
<td align="center">29</td>
|
| 251 |
+
<td align="center">Male</td>
|
| 252 |
+
<td align="center">Athens</td>
|
| 253 |
+
</tr>
|
| 254 |
+
<tr>
|
| 255 |
+
<td align="center">Olympia Simantiraki</td>
|
| 256 |
+
<td align="center">F</td>
|
| 257 |
+
<td align="center">37</td>
|
| 258 |
+
<td align="center">Female</td>
|
| 259 |
+
<td align="center">Heraklion</td>
|
| 260 |
+
</tr>
|
| 261 |
+
<tr>
|
| 262 |
+
<td align="center">Efstratios Selisios</td>
|
| 263 |
+
<td align="center">M1</td>
|
| 264 |
+
<td align="center">30</td>
|
| 265 |
+
<td align="center">Male</td>
|
| 266 |
+
<td align="center">Alexandroupoli</td>
|
| 267 |
+
</tr>
|
| 268 |
+
<tr>
|
| 269 |
+
<td align="center">Gavriela Koutsikou</td>
|
| 270 |
+
<td align="center">F1</td>
|
| 271 |
+
<td align="center">21</td>
|
| 272 |
+
<td align="center">Female</td>
|
| 273 |
+
<td align="center">Larisa</td>
|
| 274 |
+
</tr>
|
| 275 |
+
<tr>
|
| 276 |
+
<td align="center">Georgios Goupos</td>
|
| 277 |
+
<td align="center">M2</td>
|
| 278 |
+
<td align="center">22</td>
|
| 279 |
+
<td align="center">Male</td>
|
| 280 |
+
<td align="center">Larisa</td>
|
| 281 |
+
</tr>
|
| 282 |
+
<tr>
|
| 283 |
+
<td align="center">Anastasia Kechagia</td>
|
| 284 |
+
<td align="center">F2</td>
|
| 285 |
+
<td align="center">26</td>
|
| 286 |
+
<td align="center">Female</td>
|
| 287 |
+
<td align="center">Athens</td>
|
| 288 |
+
</tr>
|
| 289 |
+
</tbody>
|
| 290 |
+
</table>
|
| 291 |
+
|
| 292 |
+
</div>
|
| 293 |
+
|
| 294 |
+
<div align="center">
|
| 295 |
+
|
| 296 |
+
<strong>Table 5: Audio format specifications of the STOMA corpus recordings.</strong>
|
| 297 |
+
|
| 298 |
+
<table>
|
| 299 |
+
<thead>
|
| 300 |
+
<tr>
|
| 301 |
+
<th>Property</th>
|
| 302 |
+
<th>Value</th>
|
| 303 |
+
</tr>
|
| 304 |
+
</thead>
|
| 305 |
+
<tbody>
|
| 306 |
+
<tr>
|
| 307 |
+
<td>File format</td>
|
| 308 |
+
<td>WAV (RIFF)</td>
|
| 309 |
+
</tr>
|
| 310 |
+
<tr>
|
| 311 |
+
<td>Encoding</td>
|
| 312 |
+
<td>Pulse-Code Modulation (PCM)</td>
|
| 313 |
+
</tr>
|
| 314 |
+
<tr>
|
| 315 |
+
<td>Compression</td>
|
| 316 |
+
<td>None (uncompressed)</td>
|
| 317 |
+
</tr>
|
| 318 |
+
<tr>
|
| 319 |
+
<td>Channels</td>
|
| 320 |
+
<td>1 (mono)</td>
|
| 321 |
+
</tr>
|
| 322 |
+
<tr>
|
| 323 |
+
<td>Sampling rate</td>
|
| 324 |
+
<td>44.1 kHz</td>
|
| 325 |
+
</tr>
|
| 326 |
+
<tr>
|
| 327 |
+
<td>Bit depth</td>
|
| 328 |
+
<td>16-bit signed integer</td>
|
| 329 |
+
</tr>
|
| 330 |
+
</tbody>
|
| 331 |
+
</table>
|
| 332 |
+
|
| 333 |
+
</div>
|
| 334 |
+
|
| 335 |
+
|
| 336 |
+
## More Information
|
| 337 |
+
Check out our GitHub repository for the training and inference procedures: [STOMA: A Multi-Speaker Greek Speech Corpus](https://github.com/aangelakis/STOMA)
|