Czech MFA dictionary v2.0.0#
@techreport{mfa_czech_mfa_dictionary_2022,
author={McAuliffe, Michael and Sonderegger, Morgan},
title={Czech MFA dictionary v2.0.0},
address={\url{https://mfa-models.readthedocs.io/pronunciation dictionary/Czech/Czech MFA dictionary v2_0_0.html}},
year={2022},
month={Mar},
}
G2P models |
Installation#
Install from the MFA command line:
mfa model download dictionary czech_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 Czech transcripts.
This dictionary uses the MFA phone set for Czech, and was used in training the Czech 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 |
Glottal |
---|---|---|---|---|---|---|---|
Nasal |
Occurrences: 24,192 Examples: * čem: [tʃ ɛ m] * morf: [m o r f] * mind: [m iː n t] * mletí: [m l ɛ c iː] Occurrences: 459 Examples: * normě: [n o r m̩ ɲ ɛ] * směně: [s m̩ ɲ ɛ ɲ ɛ] * barmě: [b a r m̩ ɲ ɛ] * smrad: [s m̩ r a t] |
Occurrences: 317 Examples: * nymfa: [n ɪ ɱ f a] * nymfy: [n ɪ ɱ f ɪ] * lymfa: [l ɪ ɱ f a] * pánví: [p aː ɱ v iː] |
Occurrences: 37,543 Examples: * mind: [m iː n t] * dušan: [d u ʃ a n] * rodin: [r o ɟ ɪ n] * mýtný: [m iː t n iː] Occurrences: 23 Examples: * srnku: [s r n̩ k u] * srnka: [s r n̩ k a] * zrnko: [z r n̩ k o] * srnčí: [s r n̩ tʃ iː] |
Occurrences: 15,021 Examples: * učiní: [u tʃ ɪ ɲ iː] * síně: [s iː ɲ ɛ] * usní: [u s ɲ iː] * vnímá: [v ɲ iː m aː] |
Occurrences: 1,480 Examples: * angel: [a ŋ ɡ ɛ l] * singl: [s ɪ ŋ ɡ l̩] * lang: [l a ŋ k] * tenká: [t ɛ ŋ k aː] |
||
Stop |
Occurrences: 25,734 Examples: * trpět: [t r̩ p j ɛ t] * opéct: [o p ɛː ts t] * pouze: [p ow z ɛ] * spis: [s p ɪ s] Occurrences: 10,057 Examples: * rublu: [r u b l u] * těžby: [c ɛ ʒ b ɪ] * rybíz: [r ɪ b iː s] * bojko: [b o j k o] |
Occurrences: 36,514 Examples: * mind: [m iː n t] * trpět: [t r̩ p j ɛ t] * opéct: [o p ɛː ts t] * leto: [l ɛ t o] Occurrences: 17,967 Examples: * sady: [s a d ɪ] * jedlé: [j ɛ d l ɛː] * dušan: [d u ʃ a n] * vsedl: [f s ɛ d l̩] |
Occurrences: 5,161 Examples: * mletí: [m l ɛ c iː] * těžby: [c ɛ ʒ b ɪ] * pouť: [p ow c] * cti: [ts c ɪ] Occurrences: 3,043 Examples: * rodin: [r o ɟ ɪ n] * děsíc: [ɟ ɛ s iː ts] * sadě: [s a ɟ ɛ] * řádím: [r̝ aː ɟ iː m] |
Occurrences: 30,328 Examples: * zkusí: [s k u s iː] * kurri: [k u r ɪ] * úprku: [uː p r k u] * akcí: [a k ts iː] Occurrences: 3,643 Examples: * ehs: [ɛ ɡ ɦ] * gesto: [ɡ ɛ s t o] * glóby: [ɡ l oː b ɪ] * gajdy: [ɡ a j d ɪ] |
|||
Affricate |
Occurrences: 12,302 Examples: * výšce: [v iː ʃ ts ɛ] * opéct: [o p ɛː ts t] * ocelí: [o ts ɛ l iː] * válci: [v aː l ts ɪ] |
Occurrences: 7,864 Examples: * čem: [tʃ ɛ m] * čsú: [tʃ s uː] * učiní: [u tʃ ɪ ɲ iː] * loučí: [l ow tʃ iː] Occurrences: 122 Examples: * cage: [k ɛː dʒ] * fidži: [f ɪ dʒ ɪ] * gadžo: [ɡ a dʒ o] * džíp: [dʒ iː p] |
|||||
Sibilant |
Occurrences: 30,661 Examples: * sea: [s ɛ a] * sady: [s a d ɪ] * čsú: [tʃ s uː] * vous: [v ow s] Occurrences: 14,191 Examples: * pouze: [p ow z ɛ] * zdali: [z d a l ɪ] * czech: [ts z ɛ x] * zalít: [z a l iː t] |
Occurrences: 6,710 Examples: * výšce: [v iː ʃ ts ɛ] * dušan: [d u ʃ a n] * ušli: [u ʃ l ɪ] * rašín: [r a ʃ iː n] Occurrences: 4,221 Examples: * těžby: [c ɛ ʒ b ɪ] * žádný: [ʒ aː d n iː] * vyžít: [v ɪ ʒ iː t] * žalob: [ʒ a l o p] |
|||||
Fricative |
Occurrences: 4,998 Examples: * morf: [m o r f] * vsedl: [f s ɛ d l̩] * kavce: [k a f ts ɛ] * kufr: [k u f r̩] Occurrences: 31,756 Examples: * výšce: [v iː ʃ ts ɛ] * vous: [v ow s] * válci: [v aː l ts ɪ] * vědom: [v j ɛ d o m] |
Occurrences: 10,053 Examples: * hůl: [ɦ uː l] * háji: [ɦ aː j ɪ] * ehs: [ɛ ɡ ɦ] * prhu: [p r̩ ɦ u] |
|||||
Approximant |
Occurrences: 14,027 Examples: * jedlé: [j ɛ d l ɛː] * trpět: [t r̩ p j ɛ t] * jílů: [j iː l uː] * háji: [ɦ aː j ɪ] |
||||||
Trill |
Occurrences: 29,488 Examples: * morf: [m o r f] * rodin: [r o ɟ ɪ n] * rublu: [r u b l u] * rybíz: [r ɪ b iː s] Occurrences: 2,035 Examples: * trpět: [t r̩ p j ɛ t] * prhu: [p r̩ ɦ u] * krky: [k r̩ k ɪ] * krmí: [k r̩ m iː] |
||||||
Lateral |
Occurrences: 34,153 Examples: * jedlé: [j ɛ d l ɛː] * mletí: [m l ɛ c iː] * hůl: [ɦ uː l] * jílů: [j iː l uː] Occurrences: 891 Examples: * vsedl: [f s ɛ d l̩] * seidl: [s a j d l̩] * zvedl: [z v ɛ d l̩] * slza: [s l̩ z a] |
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: 31,065 Examples: * výšce: [v iː ʃ ts ɛ] * mind: [m iː n t] * mletí: [m l ɛ c iː] * jílů: [j iː l uː] |
Occurrences: 18,661 Examples: * dušan: [d u ʃ a n] * rublu: [r u b l u] * ušli: [u ʃ l ɪ] * zkusí: [s k u s iː] Occurrences: 4,729 Examples: * hůl: [ɦ uː l] * jílů: [j iː l uː] * čsú: [tʃ s uː] * úprku: [uː p r k u] |
|||
Occurrences: 49,143 Examples: * sady: [s a d ɪ] * háji: [ɦ aː j ɪ] * rodin: [r o ɟ ɪ n] * těžby: [c ɛ ʒ b ɪ] |
|||||
Close-Mid |
Occurrences: 56,336 Examples: * morf: [m o r f] * rodin: [r o ɟ ɪ n] * opéct: [o p ɛː ts t] * ocelí: [o ts ɛ l iː] Occurrences: 5,377 Examples: * vous: [v ow s] * pouze: [p ow z ɛ] * pouť: [p ow c] * louka: [l ow k a] Occurrences: 360 Examples: * fóre: [f oː r ɛ] * glóby: [ɡ l oː b ɪ] * chróm: [x r oː m] * fór: [f oː r] |
||||
Open-Mid |
Occurrences: 61,324 Examples: * čem: [tʃ ɛ m] * sea: [s ɛ a] * jedlé: [j ɛ d l ɛː] * výšce: [v iː ʃ ts ɛ] Occurrences: 7,796 Examples: * jedlé: [j ɛ d l ɛː] * opéct: [o p ɛː ts t] * slepé: [s l ɛ p ɛː] * bídné: [b iː d n ɛː] |
||||
Open |
Occurrences: 50,509 Examples: * sea: [s ɛ a] * sady: [s a d ɪ] * dušan: [d u ʃ a n] * seidl: [s a j d l̩] Occurrences: 18,259 Examples: * háji: [ɦ aː j ɪ] * válci: [v aː l ts ɪ] * žádný: [ʒ aː d n iː] * rámce: [r aː m ts ɛ] |
Diphthongs#
aw
ew
ow