French MFA dictionary v2.0.0#
@techreport{mfa_french_mfa_dictionary_2022,
author={McAuliffe, Michael and Sonderegger, Morgan},
title={French MFA dictionary v2.0.0},
address={\url{https://mfa-models.readthedocs.io/pronunciation dictionary/French/French MFA dictionary v2_0_0.html}},
year={2022},
month={Mar},
}
G2P models |
Installation#
Install from the MFA command line:
mfa model download dictionary french_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 French transcripts.
This dictionary uses the MFA phone set for French, and was used in training the French 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 |
Uvular |
---|---|---|---|---|---|---|---|
Nasal |
Occurrences: 48,331 Examples: * moyà: [m w a] * murés: [m y ʁ e] * mejba: [m ə ʒ b a] * asima: [a z i m a] Occurrences: 693 Examples: * simia: [s i mʲ a] * mious: [mʲ u] * mialy: [mʲ a ʎ i] * mier: [mʲ e] |
Occurrences: 41,857 Examples: * corín: [k ɔ ʁ i n] * nenny: [n ɛ ɲ i] * noyz: [n w a j e] * sen: [s e n] |
Occurrences: 13,314 Examples: * nenny: [n ɛ ɲ i] * breny: [b ʁ ɛ ɲ i] * deny: [d ə ɲ i] * buoni: [b ɥ ɔ ɲ i] |
Occurrences: 1,647 Examples: * spunk: [s p œ ŋ k] * kling: [k ʎ i ŋ ɡ] * meung: [m ø n s ɛ ŋ] * wanja: [v a ŋ ʒ a] |
|||
Stop |
Occurrences: 39,017 Examples: * campu: [k ɑ̃ p y] * aspis: [a s p i] * petry: [p ɛ t ʁ i] * popp: [p ɔ p] Occurrences: 36,484 Examples: * bartz: [b a ʁ ts] * fibi: [f i b i] * bouts: [b u] * bogo: [b ɔ ɡ o] |
Occurrences: 80,908 Examples: * titlo: [t i t l o] * soute: [s u t] * tolin: [t ɔ l ɛ̃] * petry: [p ɛ t ʁ i] Occurrences: 44,394 Examples: * riyad: [ʁ i j a d] * dutch: [d u tʃ] * hédou: [e d u] * fonde: [f ɔ̃ d] |
Occurrences: 8,296 Examples: * keyh: [c ɛ ʃ] * kiwon: [c i w ɔ̃] * kher: [c e] * mckay: [m ə c ɛ] Occurrences: 2,330 Examples: * agay: [a ɟ ɛ] * begum: [b ɛ ɟ ɔ m] * geir: [ɟ ɛ ʁ] * nagai: [n a ɟ e] |
Occurrences: 58,582 Examples: * campu: [k ɑ̃ p y] * corín: [k ɔ ʁ i n] * talk: [t a l k] * cayo: [k a j o] Occurrences: 20,991 Examples: * bogo: [b ɔ ɡ o] * graz: [ɡ ʁ ɑ] * trang: [t ʁ ɑ̃ ɡ] * bogle: [b ɔ ɡ l] |
|||
Affricate |
Occurrences: 1,200 Examples: * bartz: [b a ʁ ts] * kautz: [k o ts] * tutsi: [t y ts i] * abetz: [a b ɛ ts] |
Occurrences: 1,169 Examples: * dutch: [d u tʃ] * hutch: [u tʃ] * chuuk: [tʃ u k] * mitch: [m i tʃ] Occurrences: 660 Examples: * emoji: [e m o dʒ i] * sadj: [s a dʒ] * banjo: [b ɑ̃ dʒ o] * hadj: [a dʒ] |
|||||
Sibilant |
Occurrences: 76,149 Examples: * soute: [s u t] * aspis: [a s p i] * katio: [k a s j o] * serin: [s ə ʁ ɛ̃] Occurrences: 24,929 Examples: * jizan: [ʒ i z ɑ̃] * jasé: [ʒ ɑ z e] * abzac: [a p z a k] * asima: [a z i m a] |
Occurrences: 14,143 Examples: * eisho: [ɛ ʃ o] * chará: [ʃ a ʁ a] * shark: [ʃ a ʁ k] * shery: [ʃ ɛ ʁ i] Occurrences: 17,912 Examples: * jizan: [ʒ i z ɑ̃] * jasé: [ʒ ɑ z e] * mejba: [m ə ʒ b a] * kaija: [k e ʒ a] |
|||||
Fricative |
Occurrences: 23,971 Examples: * fibi: [f i b i] * fonde: [f ɔ̃ d] * cafés: [k a f e] * falls: [f a l] Occurrences: 25,040 Examples: * gravé: [ɡ ʁ a v e] * vala: [v a l a] * rivo: [ʁ i v o] * vaasa: [v a z a] |
Occurrences: 138,950 Examples: * bartz: [b a ʁ ts] * riyad: [ʁ i j a d] * corín: [k ɔ ʁ i n] * petry: [p ɛ t ʁ i] |
|||||
Approximant |
Occurrences: 12,100 Examples: * moyà: [m w a] * su'a: [s w a] * noyz: [n w a j e] * toua: [t w a] |
Occurrences: 34,136 Examples: * riyad: [ʁ i j a d] * cayo: [k a j o] * katio: [k a s j o] * noyz: [n w a j e] Occurrences: 3,824 Examples: * ruer: [ʁ ɥ e] * buoni: [b ɥ ɔ ɲ i] * usuel: [y z ɥ ɛ l] * ruedo: [ʁ ɥ e d o] |
|||||
Lateral |
Occurrences: 60,404 Examples: * titlo: [t i t l o] * tolin: [t ɔ l ɛ̃] * talk: [t a l k] * loup: [l u] |
Occurrences: 14,136 Examples: * liège: [ʎ ɛ ʒ] * sligo: [s ʎ i ɡ o] * lidon: [ʎ i d ɔ̃] * valië: [v a ʎ i] |
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: 109,189 Examples: * titlo: [t i t l o] * riyad: [ʁ i j a d] * jizan: [ʒ i z ɑ̃] * fibi: [f i b i] Occurrences: 29,259 Examples: * campu: [k ɑ̃ p y] * tuupa: [t u y p a] * lütte: [l y t] * inclu: [ɛ̃ k l y] |
Occurrences: 17,661 Examples: * soute: [s u t] * dutch: [d u tʃ] * loup: [l u] * tuupa: [t u y p a] |
|||
Close-Mid |
Occurrences: 73,010 Examples: * hédou: [e d u] * ruer: [ʁ ɥ e] * jasé: [ʒ ɑ z e] * noyz: [n w a j e] Occurrences: 3,667 Examples: * euris: [ø ʁ i] * meade: [m ø a d] * fœtus: [f ø t y s] * izieu: [i z j ø] |
Occurrences: 21,859 Examples: * titlo: [t i t l o] * eisho: [ɛ ʃ o] * cayo: [k a j o] * hecto: [ɛ k t o] |
|||
Occurrences: 17,260 Examples: * serin: [s ə ʁ ɛ̃] * mejba: [m ə ʒ b a] * deny: [d ə ɲ i] * rebel: [ʁ ə b ɛ l] |
|||||
Open-Mid |
Occurrences: 69,908 Examples: * petry: [p ɛ t ʁ i] * pèire: [p ɛ i ʁ] * eisho: [ɛ ʃ o] * hecto: [ɛ k t o] Occurrences: 6,111 Examples: * meurt: [m œ ʁ t] * ipetí: [a j p œ i] * spunk: [s p œ ŋ k] * törrö: [t œ ʁ œ] |
Occurrences: 61,899 Examples: * tolin: [t ɔ l ɛ̃] * corín: [k ɔ ʁ i n] * popp: [p ɔ p] * bogo: [b ɔ ɡ o] |
|||
Open |
Occurrences: 138,777 Examples: * bartz: [b a ʁ ts] * riyad: [ʁ i j a d] * aspis: [a s p i] * talk: [t a l k] |
Occurrences: 6,687 Examples: * jasé: [ʒ ɑ z e] * graz: [ɡ ʁ ɑ] * depas: [d ə p ɑ] * nâsir: [n ɑ i ʁ] |
Nasal Vowels#
Front |
Near-Front |
Central |
Near-Back |
Back |
|
---|---|---|---|---|---|
Close |
|||||
Close-Mid |
|||||
Open-Mid |
Occurrences: 15,844 Examples: * tolin: [t ɔ l ɛ̃] * inclu: [ɛ̃ k l y] * serin: [s ə ʁ ɛ̃] * yulin: [j y l ɛ̃] |
Occurrences: 23,619 Examples: * fonde: [f ɔ̃ d] * codon: [k ɔ d ɔ̃] * kiwon: [c i w ɔ̃] * thron: [t ʁ ɔ̃] |
|||
Open |
Occurrences: 36,466 Examples: * jizan: [ʒ i z ɑ̃] * campu: [k ɑ̃ p y] * trang: [t ʁ ɑ̃ ɡ] * centu: [s ɑ̃ t y] |