English (Nigeria) MFA dictionary v3.1.0#
@techreport{mfa_english_nigeria_mfa_dictionary_2024,
author={McAuliffe, Michael and Sonderegger, Morgan},
title={English (Nigeria) MFA dictionary v3.1.0},
address={\url{https://mfa-models.readthedocs.io/pronunciation dictionary/English/English (Nigeria) MFA dictionary v3_1_0.html}},
year={2024},
month={Jun},
}
|
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 v3_1_0.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: 2,637 Examples: * mass: [m a s] * palm: [pʰ a m] * aims: [e m s] * human: [uː m a n] Occurrences: 619 Examples: * minna: [mʲ i n a] * comic: [kʰ ɔ mʲ i k] * mixup: [mʲ i k s ɔ p] * emir: [ɛ mʲ i a] |
Occurrences: 9 Examples: |
Occurrences: 6,195 Examples: * want: [w ɔ n t] * skin: [s c i n] * pine: [pʰ aj n] * night: [n aj] |
Occurrences: 1,037 Examples: * tiny: [tʰ aj ɲ i] * sandy: [s a ɲ dʲ i] * ndidi: [ɲ dʲ iː dʲ i] * candy: [kʰ a ɲ dʲ i] |
Occurrences: 1,634 Examples: * think: [t̪ i ŋ k] * ewing: [j uː w i ŋ ɡ] * ancor: [a ŋ k ɔ] * young: [j ɔ ŋ ɡ] |
|||
Stop Plain |
Occurrences: 1,731 Examples: * spot: [s p ɔ t] * apron: [e p u n] * maps: [m a p s] * spar: [s p a] Occurrences: 193 Examples: * pure: [pʲ ʊ a] * speak: [s pʲ iː k] * topic: [tʰ ɔ pʲ i k] * april: [e pʲ i ɫ] Occurrences: 1,520 Examples: * broke: [b ɹ o k] * brown: [b ɹ aw n] * boils: [b ɔj ɫ z] * baby: [b e bʲ i] Occurrences: 302 Examples: * derby: [d a bʲ i] * bbc: [bʲ iː bʲ iː s iː] * shebi: [ʃ ɛ bʲ i] * beats: [bʲ iː t s] |
Occurrences: 98 Examples: * thing: [t̪ i ŋ] * think: [t̪ i ŋ k] * booth: [b u t̪] * south: [s aw t̪] Occurrences: 54 Examples: * that: [d̪ a] * the: [d̪ i] * other: [ɔ d̪ a] * dey: [d̪ e] |
Occurrences: 4,573 Examples: * adopt: [a d ɔ p t] * lasts: [l a s t s] * count: [kʰ aw n t] * lift: [ʎ i f t] Occurrences: 1,190 Examples: * steep: [s tʲ iː p] * sixty: [s i k s tʲ i] * listi: [ʎ i s tʲ i] * latin: [l a tʲ i n] Occurrences: 16 Examples: * twin: [tʷ i n] * twill: [tʷ i ɫ] * twins: [tʷ i n z] * twice: [tʷ aj s] Occurrences: 3,473 Examples: * danes: [d e n z] * cloud: [k l aw d] * goad: [ɡ o d] * drove: [d ɹ o v] Occurrences: 832 Examples: * sid: [ɛ s aj dʲ iː] * dish: [dʲ i ʃ] * dupe: [dʲ uː p] * andy: [a ɲ dʲ i] |
Occurrences: 544 Examples: * clips: [c ʎ i p s] * akin: [a c i n] * naked: [n e c i d] * skies: [s c iː z] Occurrences: 37 Examples: * queen: [cʷ iː n] * quick: [cʷ i k] * query: [cʷ i ɹ i] * quill: [cʷ i ɫ] Occurrences: 195 Examples: * ngos: [ɛ ɲ ɟ iː o z] * get: [ɟ i] * angry: [a ɲ ɟ ɹ i] * guy: [ɟ aj] Occurrences: 22 Examples: |
Occurrences: 2,156 Examples: * stuck: [s t ɔ k] * crown: [k ɹ o n] * pork: [pʰ ɔ k] * acts: [a k s] Occurrences: 129 Examples: * choir: [kʷ aj a] * akwu: [a kʷ u] * quite: [kʷ aj t] * quote: [kʷ o t] Occurrences: 1,976 Examples: * big: [bʲ i ɡ] * groom: [ɡ ɹ uː m] * lying: [l aj i ŋ ɡ] * leg: [l ɛ ɡ] Occurrences: 10 Examples: * gweje: [ɡʷ ɛ dʒ e] |
|||
Aspirated |
Occurrences: 987 Examples: * pooh: [pʰ uː] * pause: [pʰ ɔ z] * paths: [pʰ a ð z] * polls: [pʰ o ɫ z] |
Occurrences: 1,067 Examples: * tail: [tʰ e ɫ] * token: [tʰ o k ɛ n] * tuwo: [tʰ ɔ uː] * tokyo: [tʰ o cʰ i o] |
Occurrences: 314 Examples: * kings: [cʰ i ŋ ɡ z] * keg: [cʰ ɛ ɡ] * kim: [cʰ i m] * shaky: [ʃ e cʰ i] |
Occurrences: 1,351 Examples: * can't: [kʰ a n t] * karo: [kʰ a ɹ o] * okada: [o kʰ a d a] * codes: [kʰ o d z] |
||||
Affricate |
Occurrences: 690 Examples: * tries: [tʃ aj s] * chew: [tʃ i ʊ] * chili: [tʃ i ʎ i] * chase: [tʃ e s] Occurrences: 972 Examples: * joe: [dʒ o] * job: [dʒ ɔ b] * forge: [f ɔ dʒ] * jill: [dʒ i ɫ] |
|||||||
Sibilant |
Occurrences: 6,595 Examples: * cyber: [s aj b a] * stoke: [s t o k] * midst: [mʲ i d s t] * nooks: [n ʊ k s] Occurrences: 2,777 Examples: * nays: [n e z] * plans: [p l a n z] * tours: [tʰ ʊ a z] * buses: [b ɔ s ɛ z] |
Occurrences: 1,495 Examples: * sheik: [ʃ iː k] * azure: [a ʃ a] * sean: [ʃ ɔ n] * shook: [ʃ ʊ k] |
||||||
Fricative |
Occurrences: 1,318 Examples: * ralph: [ɹ a ɫ f] * golf: [ɡ ɔ ɫ f] * after: [a f t a] * faces: [f e s ɛ z] Occurrences: 349 Examples: * feels: [fʲ iː ɫ z] * fees: [fʲ iː z] * fiat: [fʲ i a t] * sofia: [s o fʲ i a] Occurrences: 1,133 Examples: * van: [v a n] * seven: [s ɛ v ɛ n] * lover: [l ɔ v a] * drove: [dʒ ɹ o v] Occurrences: 290 Examples: * envy: [ɛ m vʲ i] * navy: [n e vʲ i] * evil: [iː vʲ i ɫ] * heavy: [ç iː vʲ i] |
Occurrences: 338 Examples: * north: [n ɔ θ] * bath: [b a θ] * youth: [j uː θ] * path: [pʰ a θ] Occurrences: 159 Examples: * they: [ð e] * the: [ð e] * their: [ð ɛː] * these: [ð iː z] |
Occurrences: 124 Examples: * him: [ç i m] * hills: [ç i ɫ z] * hear: [ç i a] * heal: [ç iː ɫ] |
Occurrences: 510 Examples: * had: [h a d] * whom: [h uː m] * hut: [h ɔ t] * who's: [h uː z] |
||||
Approximant |
Occurrences: 689 Examples: * aware: [a w ɛː] * wings: [w i ŋ ɡ z] * word: [w ɔ d] * was: [w ɔ s] |
Occurrences: 4,877 Examples: * noro: [n ɔ ɹ o] * dream: [d ɹ iː m] * praat: [p ɹ a t] * opera: [ɔ p ɛ ɹ a] |
Occurrences: 312 Examples: * jung: [j ʊ ŋ ɡ] * urea: [j ʊ ɹ i a] * years: [j aː z] * yuan: [j ʊ a n] |
|||||
Lateral |
Occurrences: 1,971 Examples: * plank: [p l a ŋ k] * plum: [p l ɔ m] * slice: [s l aj s] * lands: [l a n d z] Occurrences: 1,731 Examples: * salt: [s ɔ ɫ t] * twill: [tʷ i ɫ] * hell: [h ɛ ɫ] * trial: [tʃ aj a ɫ] |
Occurrences: 1,359 Examples: * listi: [ʎ i s tʲ i] * liver: [ʎ i v a] * lever: [ʎ iː v a] * newly: [ɲ uː ʎ i] |
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: 9,562 Examples: * mary: [m ɛː ɹ i] * gifts: [ɟ i f t s] * skins: [s c i n z] * sin: [s i n] Occurrences: 1,204 Examples: * seeks: [s iː k s] * peat: [pʰ iː t] * eat: [iː t] * he's: [iː z] |
Occurrences: 707 Examples: * focus: [f o k u s] * uzoma: [u z o m a] * mosun: [m ʊ s u n] * enugu: [ɛ ɲ u ɡ uː] Occurrences: 1,036 Examples: * proof: [p ɹ uː f] * onu: [o n uː] * louis: [l uː i] * rooms: [ɹ uː m s] |
|||
Occurrences: 825 Examples: * unto: [ɔ n tʰ ʊ] * title: [tʰ aj t ʊ ɫ] * moot: [m ʊ t] * soot: [s ʊ t] |
|||||
Close-Mid |
Occurrences: 2,070 Examples: * gbure: [ɡb uː ɹ e] * jide: [dʒ i d e] * wave: [w e v] * play: [p l e] |
Occurrences: 1,482 Examples: * edo: [ɛ d o] * mando: [m a n d o] * show: [ʃ o] * notes: [n o t s] |
|||
Open-Mid |
Occurrences: 6,011 Examples: * reps: [ɹ ɛ p s] * kept: [cʰ ɛ] * enzo: [ɛ n z o] * index: [i n d ɛ k s] Occurrences: 226 Examples: * girl: [ɡ ɛː ɫ] * vary: [v ɛː ɹ i] * sarah: [s ɛː ɹ a] * wears: [w ɛː z] |
Occurrences: 16 Examples: * bury: [b ɜ ɹ i] Occurrences: 1 Examples: |
Occurrences: 6,017 Examples: * segun: [ʃ ɛ ɡ ɔ n] * golf: [ɡ ɔ f] * ross: [ɹ ɔ s] * plum: [p l ɔ m] |
||
Open |
Occurrences: 7,976 Examples: * ogaba: [o ɡ a b a] * hanya: [h a ɲ a] * ancor: [a ŋ k ɔ] * brash: [b ɹ a ʃ] Occurrences: 614 Examples: * herbs: [h aː b z] * yard: [j aː d] * kurt: [kʰ aː t] * hers: [h aː z] |
Diphthongs#
aj
aw
ɔj