Table 1 Word error rate and time-needed-to-correct of Vink-generated transcriptsWould it be possible for the table to not be in alternating white and blue rows? Furthermore the formatting is off, there should be 1 row per language, formatted the way that the row for "Arabic (Classic)" already is.
Language | Audio length (minutes) | Audio characteristics | Time-needed-to-correct (minutes) | Total words | Word Error Rate (WER) | |||
|---|---|---|---|---|---|---|---|---|
American English | 06:50 | Number of speakers | 2 | 17 | 854 | WER | 6.6% | |
Sex | F, M | Substitutions | 7 | |||||
Background noise1 | Medium | Insertions | 50 | |||||
Deletions | 0 | |||||||
Arabic (Classical Arabic) | 03:06 | Number of Speakers Sex Background noise | 1 F Low | 27.5 | 363 | WER Substitutions Insertions Deletions | 15.2% 7 20 28 | |
Bahasa Indonesia | 05:12 | Number of speakers | 2 | 10 | 465 | WER | 7.95% | |
Sex | F, F | Substitutions | 10 | |||||
Background noise | Medium | Insertions | 22 | |||||
Deletions | 5 | |||||||
Burmese | 05:05 | Number of speakers Sex Background noise | 3 M, M, F High | Transcript is nonsensical | ||||
Chinese | 05:01 | Number of speakers | 1 | 12 | 950 | WER | 0.95% | |
Sex | F | Substitutions | 8 | |||||
Background noise | Low | Insertions | 1 | |||||
Deletions | 0 | |||||||
Filipino | 5:00 | Number of speakers | 2 | 19 | 1343 | WER | 7.80% | |
Sex | F, GNB2 | Substitutions | 56 | |||||
Background noise | Medium | Insertions | 5 | |||||
Deletions | 45 | |||||||
French | 04:09 | Number of speakers | 2 | 19:57 | 611 | WER | 24% | |
Sex | F, M | Substitutions | 15 | |||||
Background noise | Medium | Insertions | 12 | |||||
Deletions | 122 | |||||||
German | 05:00 | Number of speakers | 2 | 9:40 | 676 | WER | 4.28% | |
Sex | F, F | Substitutions | 9 | |||||
Background noise | Low | Insertions | 2 | |||||
Deletions | 18 | |||||||
Malagasy | 04:41 | Number of speakers | 2 | 62 | 351 | WER | 41% | |
Sex | F, M | Substitutions | 134 | |||||
Background noise | Medium | Insertions | 12 | |||||
Deletions | 5 | |||||||
Portuguese Brazilian | 02:19 | Number of speakers | 2 | 4 | 209 | WER | 1.4% | |
Sex | F, M | Substitutions | 2 | |||||
Background noise | Medium | Insertions | 1 | |||||
Deletions | 0 | |||||||
Spanish Colombian | 06:31 | Number of speakers | 2 | 36:46 | 1111 | WER | 14.5% | |
Sex | F, F | Substitutions | 34 | |||||
Background noise | Low | Insertions | 21 | |||||
Deletions | 107 | |||||||
Tamil | 04:32 | Number of speakers | 1 | 72 | 221 | WER | 79.8% | |
Sex | M | Substitutions | 45 | |||||
Background noise | Low | Insertions | 103 | |||||
Deletions | 54 | |||||||
Turkish | 03:19 | Number of speakers | 1 | 8 | 232 | WER | 4.3% | |
Sex | F | Substitutions | 3 | |||||
Background noise | Low | Insertions | 1 | |||||
Deletions | 6 | |||||||
Yoruba | 5:56 | Number of speakers | 2 | 20 | 528 | WER | 46% | |
Sex | F, M | Substitutions | 164 | |||||
Background noise | Medium | Insertions | 36 | |||||
Deletions | 45 | |||||||