Portuguese (Brazil) MFA dictionary v2.0.0#
@techreport{mfa_portuguese_brazil_mfa_dictionary_2022,
author={McAuliffe, Michael and Sonderegger, Morgan},
title={Portuguese (Brazil) MFA dictionary v2.0.0},
address={\url{https://mfa-models.readthedocs.io/pronunciation dictionary/Portuguese/Portuguese (Brazil) MFA dictionary v2_0_0.html}},
year={2022},
month={Mar},
}
G2P models |
Installation#
Install from the MFA command line:
mfa model download dictionary portuguese_brazil_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 Portuguese transcripts.
This dictionary uses the MFA phone set for Portuguese, and was used in training the Portuguese 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 |
Alveopalatal |
Palatal |
Velar |
---|---|---|---|---|---|---|
Nasal |
Occurrences: 7,020 Examples: * emeli: [e m e ʎ i] * numa: [n u m ɐ] * mene: [m e ɲ i] * times: [tʃ i m e s] |
Occurrences: 4,073 Examples: * dona: [d o n ɐ] * numa: [n u m ɐ] * noeli: [n o e ʎ i] * canal: [k a n a w] |
Occurrences: 1,513 Examples: * mene: [m e ɲ i] * helen: [e l e ɲ i] * alane: [a l ɐ ɲ i] * one: [o ɲ i] |
|||
Stop |
Occurrences: 7,104 Examples: * aporá: [a p o ɾ a] * space: [s p e j s i] * frapê: [f ɾ a p e] * tripa: [t ɾ i p ɐ] Occurrences: 4,090 Examples: * bote: [b ɔ tʃ i] * darub: [d a ɾ u b] * rabos: [x a b u s] * ebó: [e b ɔ] |
Occurrences: 10,391 Examples: * tenso: [t ẽ s u] * tripa: [t ɾ i p ɐ] * tam: [t ɐ̃ w̃] * steve: [s t e v i] Occurrences: 8,733 Examples: * dona: [d o n ɐ] * darub: [d a ɾ u b] * mode: [m o d e] * depor: [d e p o ɾ] |
Occurrences: 1,059 Examples: * toque: [t ɔ c i] * que: [c i] * quedo: [c e d u] * saque: [s a c i] Occurrences: 407 Examples: * higgs: [i ʒ e ɟ e ɛ s i] * bug: [b u ɟ i] * gate: [ɟ e j tʃ] * ongs: [õ ɟ i s] |
Occurrences: 8,720 Examples: * curas: [k u ɾ ɐ s] * cuspi: [k u s p i] * acaso: [a k a z u] * lucas: [l u k ɐ s] Occurrences: 2,970 Examples: * grime: [ɡ ɾ i m i] * gol: [ɡ o w] * cemig: [s e m i ɡ] * figos: [f i ɡ u s] |
||
Affricate |
Occurrences: 4,018 Examples: * bote: [b ɔ tʃ i] * times: [tʃ i m e s] * irati: [i ɾ a tʃ i] * haiti: [a i tʃ i] Occurrences: 2,466 Examples: * audi: [a w dʒ i] * pedi: [p e dʒ i] * admin: [a dʒ m ĩ] * apodi: [a p o dʒ i] |
|||||
Sibilant |
Occurrences: 19,383 Examples: * tenso: [t ẽ s u] * space: [s p e j s i] * curas: [k u ɾ ɐ s] * cuspi: [k u s p i] Occurrences: 3,464 Examples: * ousou: [o w z o w] * acaso: [a k a z u] * raiza: [x a i z ɐ] * fazer: [f a z e x] |
Occurrences: 1,777 Examples: * jaez: [ʒ ɐ e ʃ] * xhosa: [ʃ ɔ z ɐ] * chove: [ʃ o v e] * uchoa: [u ʃ o ɐ] Occurrences: 1,981 Examples: * higgs: [i ʒ e ɟ e ɛ s i] * jaez: [ʒ ɐ e ʃ] * jurou: [ʒ u ɾ o w] * joe: [ʒ o e] |
||||
Fricative |
Occurrences: 3,144 Examples: * frapê: [f ɾ a p e] * fazer: [f a z e x] * filé: [f i l ɛ] * ficap: [f i k a p e] Occurrences: 4,157 Examples: * river: [x i v e x] * steve: [s t e v i] * woo: [v o u] * névoa: [n ɛ v w ɐ] |
|||||
Approximant |
Occurrences: 4,893 Examples: * ousou: [o w z o w] * audi: [a w dʒ i] * armou: [a x m o w] * seu: [s e w] Occurrences: 2,215 Examples: * tam: [t ɐ̃ w̃] * eram: [e ɾ ɐ̃ w̃] * pirão: [p i ɾ ɐ̃ w̃] * vazão: [v a z ɐ̃ w̃] |
Occurrences: 4,526 Examples: * space: [s p e j s i] * haiti: [a j tʃ i] * aurea: [a w ɾ j ɐ] * seis: [s e j s] Occurrences: 1,789 Examples: * detém: [d e t ẽ j̃] * têm: [t ẽ j̃] * expõe: [i s p õ j̃] * ximen: [ʃ i m ẽ j̃] |
||||
Tap |
Occurrences: 12,843 Examples: * aporá: [a p o ɾ a] * darub: [d a ɾ u b] * frapê: [f ɾ a p e] * curas: [k u ɾ ɐ s] |
|||||
Lateral |
Occurrences: 4,606 Examples: * helen: [e l e ɲ i] * lopes: [l ɔ p i s] * platô: [p l a t o] * lucas: [l u k ɐ s] |
Occurrences: 2,918 Examples: * emeli: [e m e ʎ i] * noeli: [n o e ʎ i] * gales: [ɡ a ʎ i s] * still: [s t i ʎ a l] |
Vowels#
Vowel symbols to the left of are unrounded and those to the right are rounded.
Oral Vowels#
Front |
Near-Front |
Central |
Near-Back |
Back |
|
---|---|---|---|---|---|
Close |
Occurrences: 23,624 Examples: * bote: [b ɔ tʃ i] * river: [x i v e x] * emeli: [e m e ʎ i] * mene: [m e ɲ i] |
Occurrences: 14,236 Examples: * numa: [n u m ɐ] * tenso: [t ẽ s u] * darub: [d a ɾ u b] * curas: [k u ɾ ɐ s] |
|||
Close-Mid |
Occurrences: 14,248 Examples: * river: [x i v e x] * emeli: [e m e ʎ i] * mene: [m e ɲ i] * space: [s p e j s i] |
Occurrences: 9,534 Examples: * aporá: [a p o ɾ a] * dona: [d o n ɐ] * ousou: [o w z o w] * armou: [a x m o w] |
|||
Open-Mid |
Occurrences: 3,163 Examples: * higgs: [i ʒ e ɟ e ɛ s i] * névoa: [n ɛ v w ɐ] * sexos: [s ɛ k s u s] * sopé: [s o p ɛ] |
Occurrences: 2,094 Examples: * bote: [b ɔ tʃ i] * lopes: [l ɔ p i s] * iperó: [i p i ɾ ɔ] * ebó: [e b ɔ] |
|||
Occurrences: 9,897 Examples: * dona: [d o n ɐ] * numa: [n u m ɐ] * curas: [k u ɾ ɐ s] * tripa: [t ɾ i p ɐ] |
|||||
Open |
Occurrences: 24,142 Examples: * aporá: [a p o ɾ a] * darub: [d a ɾ u b] * frapê: [f ɾ a p e] * audi: [a w dʒ i] |
Nasal Vowels#
Front |
Near-Front |
Central |
Near-Back |
Back |
|
---|---|---|---|---|---|
Close |
Occurrences: 2,800 Examples: * admin: [a dʒ m ĩ] * imbé: [ĩ b ɛ] * ninfa: [ɲ ĩ f ɐ] * setim: [s e tʃ ĩ] |
Occurrences: 534 Examples: * boom: [b ũ] * uns: [ũ s] * bunda: [b ũ d ɐ] * ruins: [x ũ j̃ s] |
|||
Close-Mid |
Occurrences: 4,026 Examples: * tenso: [t ẽ s u] * lenda: [l ẽ d a] * detém: [d e t ẽ j̃] * têm: [t ẽ j̃] |
Occurrences: 2,429 Examples: * éxons: [ɛ k s õ s] * ron: [x õ] * sólon: [s ɔ l õ] * ronda: [x õ d ɐ] |
|||
Open-Mid |
|||||
Occurrences: 5,483 Examples: * tam: [t ɐ̃ w̃] * eram: [e ɾ ɐ̃ w̃] * dando: [d ɐ̃ d u] * pirão: [p i ɾ ɐ̃ w̃] |
|||||
Open |