English (Nigeria) MFA dictionary v2.2.1#
@techreport{mfa_english_nigeria_mfa_dictionary_2023,
author={McAuliffe, Michael and Sonderegger, Morgan},
title={English (Nigeria) MFA dictionary v2.2.1},
address={\url{https://mfa-models.readthedocs.io/pronunciation dictionary/English/English (Nigeria) MFA dictionary v2_2_1.html}},
year={2023},
month={May},
}
G2P models |
Installation#
Install from the MFA command line:
mfa model download dictionary english_nigeria_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](https://raw.githubusercontent.com/MontrealCorpusTools/mfa-models/main/dictionary/english/mfa/English (Nigeria) MFA dictionary v2_2_1.dict).
Intended use#
This dictionary is intended for forced alignment of English transcripts.
This dictionary uses the MFA phone set for English, and was used in training the English 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 |
Dental |
Alveolar |
Alveopalatal |
Palatal |
Velar |
Glottal |
---|---|---|---|---|---|---|---|---|
Nasal |
Occurrences: 11,448 Examples: * modem: [m o d ɛ m] * foism: [f o i z a m] * mayst: [m e s t] * mega: [m ɛ ɡ a] Occurrences: 2,704 Examples: * mere: [mʲ i a] * semi: [s ɛ mʲ i] * mixin: [mʲ i k s i n] * mealy: [mʲ iː ʎ i] |
Occurrences: 24 Examples: |
Occurrences: 23,067 Examples: * neddy: [n ɛ dʲ i] * jawan: [dʒ a w a n] * bench: [b ɛ n tʃ] * mane: [m e n] |
Occurrences: 4,914 Examples: * gnu: [ɲ uː] * panic: [pʰ a ɲ i k] * ionic: [aj ɔ ɲ i k] * nipa: [ɲ iː pʰ a] |
Occurrences: 4,934 Examples: * skunk: [s k ɔ ŋ k] * ungot: [ɔ ŋ ɡ ɔ t] * ging: [ɟ i ŋ ɡ] * axing: [a k s i ŋ ɡ] |
|||
Stop |
Occurrences: 6,279 Examples: * rupel: [ɹ uː p ɛ ɫ] * pshaw: [p ʃ ɔ] * cap: [kʰ a p] * coupe: [kʰ uː p] Occurrences: 7,636 Examples: * plumb: [p l ɔ m b] * bury: [b ɛ ɹ i] * biota: [b aj o t a] * nobly: [n o b ʎ i] |
Occurrences: 47 Examples: Occurrences: 53 Examples: |
Occurrences: 16,779 Examples: * spilt: [s pʲ i ɫ t] * track: [t ɹ a k] * stoep: [s t uː p] * it's: [i t s] Occurrences: 12,982 Examples: * mould: [m o ɫ d] * shide: [ʃ aj d] * homed: [h o m d] * defog: [d ɛ f ɔ ɡ] |
Occurrences: 2,530 Examples: * skew: [s c uː] * likin: [ʎ iː c i n] * gskew: [dʒ iː s c uː] * skein: [s c e n] Occurrences: 1,046 Examples: * ginn: [ɟ i n] * gleby: [ɟ ʎ iː bʲ i] * bigly: [bʲ i ɟ ʎ i] * glebe: [ɟ ʎ iː b] |
Occurrences: 10,202 Examples: * pacts: [pʰ a k s] * close: [k l o s] * gleek: [ɟ ʎ iː k] * krait: [k ɹ aj t] Occurrences: 7,728 Examples: * gaze: [ɡ e z] * hagar: [ç e ɡ a] * pego: [pʰ iː ɡ o] * sing: [s i ŋ ɡ] |
Occurrences: 136 Examples: * but: [b a ʔ] * apart: [a pʰ a ʔ] * mate: [m e ʔ] * suit: [s j uː ʔ] |
||
Affricate |
Occurrences: 2,396 Examples: * chalk: [tʃ ɔ k] * belch: [b ɛ ɫ tʃ] * chow: [tʃ aw] * chock: [tʃ ɔ k] Occurrences: 3,827 Examples: * sigil: [s i dʒ i ɫ] * junco: [dʒ ɔ ŋ kʰ o] * ajar: [a dʒ a] * zhu: [dʒ uː] |
|||||||
Sibilant |
Occurrences: 25,397 Examples: * cyber: [s aj b a] * dulse: [d ɔ ɫ s] * dusts: [d ɔ s t s] * sfnal: [ɛ s ɛ f n a ɫ] Occurrences: 9,030 Examples: * zis: [z i s] * bags: [b a ɡ z] * booms: [b uː m z] * abysm: [a bʲ i z a m] |
Occurrences: 5,490 Examples: * shift: [ʃ i f t] * shrek: [ʃ ɹ ɛ k] * abash: [a b a ʃ] * apish: [e pʲ i ʃ] |
||||||
Fricative |
Occurrences: 5,769 Examples: * flary: [f l ɛː ɹ i] * frock: [f ɹ ɔ k] * farse: [f a s] * fussy: [f ɔ s i] Occurrences: 1,603 Examples: * filly: [fʲ i ʎ i] * raphe: [ɹ e fʲ i] * taffy: [tʰ a fʲ i] * fib: [fʲ i b] Occurrences: 3,930 Examples: * wives: [w aj v z] * arvo: [a v o] * vide: [v aj d] * vent: [v ɛ n t] Occurrences: 1,147 Examples: * veep: [vʲ iː p] * via: [vʲ i a] * civic: [s i vʲ i k] * villa: [vʲ i l a] |
Occurrences: 1,811 Examples: * ethic: [ɛ θ i k] * forth: [f ɔ θ] * thane: [θ e n] * meth: [m ɛ θ] Occurrences: 449 Examples: * then: [ð ɛ n] * thus: [ð ɔ s] * adhan: [a ð a n] * dalet: [d a ʎ i ð] |
Occurrences: 695 Examples: * heals: [ç iː ɫ z] * hinny: [ç i ɲ i] * humes: [ç uː m z] * hitt: [ç i t] |
Occurrences: 2,413 Examples: * hurst: [h aː s t] * mhm: [m h m] * hyawa: [h aj a w a] * hum: [h ɔ m] |
||||
Approximant |
Occurrences: 2,574 Examples: * warm: [w ɔ m] * weber: [w ɛ b a] * waspi: [w ɔ s pʲ i] * sweat: [s w ɛ t] |
Occurrences: 19,074 Examples: * brink: [b ɹ i ŋ k] * read: [ɹ iː d] * idris: [i d ɹ i s] * rangy: [ɹ e n dʒ i] |
Occurrences: 1,210 Examples: * yexes: [j ɛ k s ɛ z] * uley: [j uː ʎ i] * units: [j uː ɲ i t s] * ukie: [j uː cʰ i] |
|||||
Lateral |
Occurrences: 9,222 Examples: * julep: [dʒ uː l ɛ p] * lapel: [l a pʰ ɛ ɫ] * loss: [l ɔ s] * isolo: [i s o l o] Occurrences: 7,424 Examples: * sold: [s o ɫ d] * dial: [d aj a ɫ] * gilt: [ɟ i ɫ t] * mils: [mʲ i ɫ z] |
Occurrences: 5,852 Examples: * dimly: [dʲ i m ʎ i] * melee: [m ɛ ʎ i] * plea: [p ʎ iː] * leans: [ʎ iː n z] |
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: 40,495 Examples: * myths: [mʲ i θ s] * warly: [w ɔ ʎ i] * krym: [c ɹ i m] * qubit: [cʰ uː bʲ i t] Occurrences: 6,252 Examples: * chi: [tʃ iː] * lea: [ʎ iː] * thede: [θ iː d] * weed: [w iː d] |
Occurrences: 4,060 Examples: * aaron: [a u n] * lion: [l aj u n] * jotun: [j o t u n] * kunes: [cʰ u n z] Occurrences: 4,741 Examples: * hew: [ç uː] * udo: [uː d o] * undo: [ɔ n d uː] * fufu: [f uː f uː] |
|||
Occurrences: 3,808 Examples: * gold: [ɡ ɔ ʊ ɫ d] * book: [b ʊ k] * umaru: [ʊ m a ɹ uː] * lured: [ʎ ʊ a d] |
|||||
Close-Mid |
Occurrences: 7,827 Examples: * haig: [ç e ɡ] * kane: [cʰ e n] * clave: [k l e v] * blake: [b l e k] |
Occurrences: 7,119 Examples: * clio: [c ʎ iː o] * oakum: [o k u m] * moro: [m ɔ ɹ o] * kola: [kʰ o l a] |
|||
Open-Mid |
Occurrences: 21,185 Examples: * recce: [ɹ ɛ cʰ i] * rebec: [ɹ ɛ b ɛ k] * sid: [ɛ s aj dʲ iː] * faq: [ɛ f e cʰ uː] Occurrences: 761 Examples: * erie: [ɛː ɹ i] * kerr: [cʰ ɛː] * glary: [ɡ l ɛː ɹ i] * laird: [l ɛː d] |
Occurrences: 136 Examples: * viera: [vʲ iː ɜ ɹ a] * durst: [d ɜ s t] * circe: [s ɜ s i] * wyrd: [w ɜ d] Occurrences: 3 Examples: |
Occurrences: 24,034 Examples: * hucks: [h ɔ k s] * fox: [f ɔ k s] * orfen: [ɔ f ɛ n] * gourd: [ɡ ɔ d] |
||
Open |
Occurrences: 35,590 Examples: * myna: [m aj n a] * sat: [s a] * clans: [k l a n z] * pusta: [pʰ ʊ s t a] Occurrences: 2,167 Examples: * serf: [s aː f] * merch: [m aː tʃ] * quirk: [kʷ aː k] * birth: [b aː ð] |
Diphthongs#
aj
aw
ɔj