Vietnamese MFA dictionary v2.0.0#
@techreport{mfa_vietnamese_mfa_dictionary_2022,
author={McAuliffe, Michael and Sonderegger, Morgan},
title={Vietnamese MFA dictionary v2.0.0},
address={\url{https://mfa-models.readthedocs.io/pronunciation dictionary/Vietnamese/Vietnamese MFA dictionary v2_0_0.html}},
year={2022},
month={Mar},
}
Acoustic models |
Installation#
Install from the MFA command line:
mfa model download dictionary vietnamese_mfa
Or download from the release page.
The dictionary available from the release page and command line installation has pronunciation and silence probabilities estimated as part acoustic model training (see Silence probability format and training pronunciation probabilities for more information. If you would like to use the version of this dictionary without probabilities, please see the plain dictionary.
Intended use#
This dictionary is intended for forced alignment of Vietnamese transcripts.
This dictionary uses the MFA phone set for Vietnamese, and was used in training the Vietnamese MFA acoustic model. Pronunciations can be added on top of the dictionary, as long as no additional phones are introduced.
Performance Factors#
When trying to get better alignment accuracy, adding pronunciations is generally helpful, especially for different styles and dialects. The most impactful improvements will generally be seen when adding reduced variants that involve deleting segments/syllables common in spontaneous speech. Alignment must include all phones specified in the pronunciation of a word, and each phone has a minimum duration (by default 10ms). If a speaker pronounces a multisyllabic word with just a single syllable, it can be hard for MFA to fit all the segments in, so it will lead to alignment errors on adjacent words as well.
Ethical considerations#
Deploying any Speech-to-Text model into any production setting has ethical implications. You should consider these implications before use.
Demographic Bias#
You should assume every machine learning model has demographic bias unless proven otherwise. For pronunciation dictionaries, it is often the case that transcription accuracy and lexicon coverage for the prestige variety modeled in this dictionary compared to other variants. If you are using this dictionary in production, you should acknowledge this as a potential issue.
IPA Charts#
Consonants#
Obstruent symbols to the left of are unvoiced and those to the right are voiced.
Manner |
Labial |
Labiodental |
Alveolar |
Retroflex |
Palatal |
Velar |
Glottal |
---|---|---|---|---|---|---|---|
Nasal |
Occurrences: 1,838 Examples: * chôm: [c oː˨˥ m] * nam: [n aː˨˥ m] * cám: [k aː˦˥ m] * mắn: [m a˦˥ ŋ] |
Occurrences: 1,630 Examples: * nam: [n aː˨˥ m] * thính: [tʰ ɨ˦˥ n] * băn: [ɓ a˨˨ n] * viền: [v iə˨˨ n] |
Occurrences: 664 Examples: * thính: [tʰ i˨˦ ɲ] * thịnh: [tʰ i˨˩ˀ ɲ] * nho: [ɲ ɔː˨˥] * nhờ: [ɲ əː˦˨] |
Occurrences: 2,753 Examples: * ngạt: [ŋ aː˨˩ k] * ngoi: [ŋ ɔː˨˥ j] * ngứa: [ŋ ɨə˦˥] * dương: [j ɨː˨˥ ŋ] |
|||
Stop Plain |
Occurrences: 367 Examples: * nượp: [n ɨə˨˩ p] * lép: [l ɛː˦˥ p] * giáp: [j aː˦˥ p] * hợp: [h əː˨˩ p] |
Occurrences: 1,332 Examples: * ngạt: [ŋ aː˨˩ t] * dứt: [z ɨ˦˥ t] * bạt: [ɓ aː˨˩ t] * tắc: [t a˦˥ k] |
Occurrences: 303 Examples: * trườn: [ʈ ɨː˦˨ ŋ] * tro: [ʈ ɔː˨˥] * trứng: [ʈ ɨ˦˥ ŋ] * trĩu: [ʈ i˨˨ˀ w] |
Occurrences: 518 Examples: * chôm: [c oː˨˥ m] * chiết: [c iː˦˥ k] * chớm: [c əː˦˥ m] * chia: [c iə˨˥] |
Occurrences: 1,619 Examples: * quý: [k w iː˨˦] * ngạt: [ŋ aː˨˩ k] * cám: [k aː˦˥ m] * sắc: [s a˦˥ k] Occurrences: 250 Examples: * gờm: [ɡ əː˦˨ m] * gác: [ɡ aː˦˥ k] * gồng: [ɡ ɔ˦˨ ŋm] * gục: [ɡ u˨˩ kp] |
Occurrences: 493 Examples: * ổng: [ʔ o˨˩˨ ŋm] * ướt: [ʔ ɨə˦˥ t] * ươi: [ʔ ɨə˨˨ j] * yếu: [ʔ iə˨˦ w] |
|
Implosive |
Occurrences: 720 Examples: * băn: [ɓ a˨˥ ŋ] * bạt: [ɓ aː˨˩ k] * bái: [ɓ aː˦˥ j] * bốt: [ɓ oː˦˥ t] |
Occurrences: 693 Examples: * đều: [ɗ eː˦˨ w] * đoá: [ɗ w aː˦˥] * đảo: [ɗ aː˨˨ˀ w] * đang: [ɗ aː˨˥ ŋ] |
|||||
Affricate |
Occurrences: 392 Examples: * chôm: [tɕ oː˨˨ m] * chiết: [tɕ iə˦˥ t] * chớm: [tɕ əː˨˦ m] * chia: [tɕ iə˨˨] |
||||||
Sibilant |
Occurrences: 881 Examples: * sắc: [s a˦˥ k] * xuất: [s w ə˦˥ k] * xém: [s ɛː˦˥ m] * sang: [s aː˨˥ ŋ] Occurrences: 501 Examples: * dương: [z ɨə˨˨ ŋ] * dứt: [z ɨ˦˥ t] * răng: [z a˨˨ ŋ] * rưới: [z ɨə˨˦ j] |
||||||
Fricative |
Occurrences: 323 Examples: * phác: [f aː˦˥ k] * phan: [f aː˨˥ ŋ] * phình: [f i˨˨ ɲ] * phễu: [f eː˨ˀ˥ w] Occurrences: 516 Examples: * viền: [v iə˨˨ n] * vèo: [v ɛː˦˨ w] * vờn: [v əː˦˨ ŋ] * vọng: [v ɔ˨˩ ŋm] |
Occurrences: 696 Examples: * hãi: [h aː˨ˀ˥ j] * hôi: [h oː˨˥ j] * hít: [h iː˦˥ t] * huyên: [h w iə˨˨ n] |
|||||
Approximant |
Occurrences: 2,183 Examples: * quý: [k w iː˨˦] * lầu: [l ə˦˨ w] * đều: [ɗ eː˦˨ w] * khuy: [x w iː˨˨] |
Occurrences: 2,284 Examples: * lười: [l ɨə˨˨ j] * ngoi: [ŋ ɔː˨˥ j] * dương: [j ɨː˨˥ ŋ] * dứt: [j ɨ˦˥ k] |
|||||
Trill |
Occurrences: 385 Examples: * răng: [r a˨˥ ŋ] * rưới: [r ɨː˦˥ j] * rái: [r aː˦˥ j] * run: [r u˨˥ ŋm] |
||||||
Lateral |
Occurrences: 797 Examples: * lười: [l ɨə˨˨ j] * lầu: [l ə˦˨ w] * lào: [l aː˦˨ w] * liêng: [l iə˨˨ ŋ] |
Vowels#
Vowel symbols to the left of are unrounded and those to the right are rounded.
Front |
Near-Front |
Central |
Near-Back |
Back |
|
---|---|---|---|---|---|
Close |
Occurrences: 531 Examples: * thi: [tʰ i˨˨] * tính: [t i˨˦ ɲ] * lĩnh: [l i˨ˀ˥ ɲ] * phiu: [f i˨˥ w] Occurrences: 802 Examples: * hít: [h iː˦˥ t] * xịn: [s iː˨˩ˀ n] * bỉm: [ɓ iː˨˩˨ m] * trĩu: [tɕ iː˨ˀ˥ w] |
Occurrences: 760 Examples: * sững: [s ɨ˨ˀ˥ ŋ] * thính: [tʰ ɨ˨ˀ˦ n] * cùm: [k ɨ˨˩ m] * min: [m ɨ˨˨ n] Occurrences: 345 Examples: * vượt: [v ɨː˨˩˨ k] * giữ: [z ɨː˨ˀ˥] * ngữ: [ŋ ɨː˨ˀ˥] * ngự: [ŋ ɨː˨˩ˀ] |
Occurrences: 695 Examples: * đụng: [ɗ u˨˩ˀ ŋm] * thúy: [tʰ u˨ˀ˦ j] * dũng: [z u˨ˀ˥ ŋm] * run: [r u˨˥ ŋm] Occurrences: 446 Examples: * bụt: [ɓ uː˨˩ t] * muỗng: [m uː˨˩˦ ŋ] * uổng: [ʔ uː˨˩˦ ŋ] * túi: [t uː˨˦ j] |
||
Close-Mid |
Occurrences: 38 Examples: * lệnh: [l e˨˩ˀ ɲ] * tếch: [t e˦˥ c] * lênh: [l e˨˨ ɲ] * vênh: [v e˨˨ ɲ] Occurrences: 396 Examples: * lều: [l eː˦˨ w] * huệ: [h w eː˨˩ˀ] * quê: [w eː˨˥] * rệt: [z eː˨˩ t] |
Occurrences: 239 Examples: * lột: [l o˨˩˨ k] * bỗng: [ɓ o˨ˀ˥ ŋm] * ngốc: [ŋ o˦˥ kp] * khổng: [x o˨˩˨ ŋm] Occurrences: 592 Examples: * hôi: [h oː˨˨ j] * hồi: [h oː˦˨ j] * chồm: [c oː˦˨ m] * phố: [f oː˨ˀ˦] |
|||
Occurrences: 1,164 Examples: * dâm: [j ə˨˥ m] * thân: [tʰ ə˨˥ ŋ] * rấm: [r ə˨ˀ˦ m] * trầm: [ʈ ə˦˨ m] Occurrences: 832 Examples: * tếch: [t əː˦˥ t] * xếch: [s əː˦˥ t] * giỡn: [z əː˨ˀ˥ n] * hởi: [h əː˨˨ˀ j] |
|||||
Open-Mid |
Occurrences: 816 Examples: * tem: [t ɛː˨˥ m] * bẹn: [ɓ ɛː˨˩ˀ n] * nhè: [ɲ ɛː˦˨] * kẻng: [k ɛː˨˩˦ ŋ] |
Occurrences: 447 Examples: * gióng: [z ɔ˨˦ ŋm] * ổng: [ʔ ɔ˨˨ˀ ŋm] * lọng: [l ɔ˨˩ˀ ŋm] * óng: [ʔ ɔ˨ˀ˦ ŋm] Occurrences: 736 Examples: * cõi: [k ɔː˨ˀ˥ j] * nõn: [n ɔː˨ˀ˥ n] * phọt: [f ɔː˨˩ k] * khóm: [x ɔː˦˥ m] |
|||
Open |
Occurrences: 1,280 Examples: * rãnh: [z a˨ˀ˥ ɲ] * nắm: [n a˨˦ m] * quằn: [w a˦˨ ŋ] * nhảy: [ɲ a˨˩˦ j] Occurrences: 2,147 Examples: * rái: [r aː˦˥ j] * ngạt: [ŋ aː˨˩ k] * đoá: [ɗ w aː˨˦] * trại: [tɕ aː˨˩ˀ j] |
Diphthongs#
iə
uə
ɨə
Tones#
˦˥
˦˨
˨ˀ˥
˨ˀ˦
˨˥
˨˦
˨˨
˨˨ˀ
˨˩
˨˩ˀ
˨˩˦
˨˩˨