Russian CV dictionary v2.0.0#
@misc{Ahn_Chodroff_2022,
author={Ahn, Emily and Chodroff, Eleanor},
title={VoxCommunis Corpus},
address={\url{https://osf.io/t957v}},
publisher={OSF},
year={2022},
month={Jan}
}
Acoustic models |
Installation#
Install from the MFA command line:
mfa model download dictionary russian_cv
Or download from the release page.
Intended use#
This dictionary is intended for forced alignment of Russian transcripts.
This dictionary uses the Epitran phone set for Russian, and was used in training the Russian Epitran 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 |
Retroflex |
Palatal |
Velar |
Glottal |
---|---|---|---|---|---|---|---|---|
Nasal |
Occurrences: 10,414 Examples: * мраке: [m r a kʲ e] * эмма: [e m m a] * чтим: [t͡ɕʲ tʲ i m] * всем: [v sʲ e m] Occurrences: 4,745 Examples: * мере.: [mʲ e rʲ e] * меж: [mʲ e ʒ] * нами,: [n a mʲ i] * мер?: [mʲ e r] |
Occurrences: 30,527 Examples: * кейн.: [kʲ e j n] * ну…: [n u] * фонд.: [f o n d] * несет: [n e sʲ e t] Occurrences: 200 Examples: * рознь: [r o z nʲ] * тень.: [tʲ e nʲ] * дрянь: [d rʲ a nʲ] * казнь: [k a z nʲ] |
||||||
Stop |
Occurrences: 12,523 Examples: * пробу: [p r o b u] * кипр: [kʲ i p r] * праву: [p r a v u] * слепа: [s lʲ e p a] Occurrences: 4,983 Examples: * пробу: [p r o b u] * зубы: [z u b ɨ] * буди: [b u dʲ i] * букве: [b u k vʲ e] |
Occurrences: 17,081 Examples: * несет: [n e sʲ e t] * тон.: [t o n] * ритм,: [rʲ i t m] * лифт: [lʲ i f t] Occurrences: 8,334 Examples: * фонд.: [f o n d] * долгу: [d o l ɡ u] * дочь: [d o t͡ɕʲ] * дрожь: [d r o ʒ] |
Occurrences: 4 Examples: |
Occurrences: 11,265 Examples: * кожи: [k o ʒ i] * букве: [b u k vʲ e] * коли: [k o lʲ i] * банка: [b a n k a] Occurrences: 6,871 Examples: * долгу: [d o l ɡ u] * года: [ɡ o d a] * погод: [p o ɡ o d] * годом: [ɡ o d o m] |
||||
Affricate |
Occurrences: 2,696 Examples: * цел: [t͡s e l] * лицо,: [lʲ i t͡s o] * ницше: [n i t͡s ʂ e] * цифры: [t͡s i f r ɨ] |
Occurrences: 110 Examples: * джон: [d͡ʒ o n] * джо: [d͡ʒ o] * джим.: [d͡ʒ i m] * бундж: [b u n d͡ʒ] |
||||||
Sibilant |
Occurrences: 19,709 Examples: * сашки: [s a ʂ kʲ i] * сучья: [s u t͡ɕʲ a] * леса: [lʲ e s a] * риса,: [rʲ i s a] Occurrences: 5,822 Examples: * несет: [n e sʲ e t] * всем: [v sʲ e m] * сеня.: [sʲ e n a] * сил: [sʲ i l] Occurrences: 6,895 Examples: * зубы: [z u b ɨ] * взор: [v z o r] * назад: [n a z a d] * зам: [z a m] Occurrences: 864 Examples: * визит: [vʲ i zʲ i t] * взяв: [v zʲ a v] * лазил: [l a zʲ i l] * зимам: [zʲ i m a m] |
Occurrences: 3,801 Examples: * меж: [mʲ e ʒ] * кожи: [k o ʒ i] * дрожь: [d r o ʒ] * уж,: [u ʒ] |
Occurrences: 3,154 Examples: * сашки: [s a ʂ kʲ i] * шею.: [ʂ e u] * лучше: [l u t͡ɕʲ ʂ e] * чашки: [t͡ɕʲ a ʂ kʲ i] Occurrences: 204 Examples: * учишь: [u t͡ɕʲ i ʂʲː] * вишь,: [vʲ i ʂʲː] * тушь.: [t u ʂʲː] * рвешь: [r vʲ e ʂʲː] |
Occurrences: 2,156 Examples: * леща.: [lʲ e ɕː a] * ищете: [i ɕː e tʲ e] * ящик.: [j a ɕː i k] * чаще,: [t͡ɕʲ a ɕː e] Occurrences: 9 Examples: * вещь: [vʲ e ɕːʲ] * мощь: [m o ɕːʲ] * вещь.: [vʲ e ɕːʲ] * мощь,: [m o ɕːʲ] |
||||
Fricative |
Occurrences: 892 Examples: * фонд.: [f o n d] * лифт: [lʲ i f t] * шкаф?: [ʂ k a f] * цифры: [t͡s i f r ɨ] Occurrences: 656 Examples: * сфере: [s fʲ e rʲ e] * фин: [fʲ i n] * фильм: [fʲ i lʲ m] * сфера: [s fʲ e r a] Occurrences: 14,541 Examples: * праву: [p r a v u] * взор: [v z o r] * ввиду: [v vʲ i d u] * всем: [v sʲ e m] Occurrences: 4,691 Examples: * букве: [b u k vʲ e] * ввиду: [v vʲ i d u] * весах: [vʲ e s a x] * ввел: [v vʲ e l] |
Occurrences: 2 Examples: |
||||||
Approximant |
Occurrences: 6,332 Examples: * кейн.: [kʲ e j n] * ясно: [j a s n o] * свой,: [s v o j] * ярмо: [j a r m o] |
|||||||
Trill |
Occurrences: 17,322 Examples: * пробу: [p r o b u] * мраке: [m r a kʲ e] * кипр: [kʲ i p r] * дрожь: [d r o ʒ] Occurrences: 8,715 Examples: * мере.: [mʲ e rʲ e] * теряя: [tʲ e rʲ a a] * риса,: [rʲ i s a] * ритм,: [rʲ i t m] |
|||||||
Lateral |
Occurrences: 9,079 Examples: * долгу: [d o l ɡ u] * лучше: [l u t͡ɕʲ ʂ e] * цел: [t͡s e l] * кидал: [kʲ i d a l] Occurrences: 10,642 Examples: * леща.: [lʲ e ɕː a] * леса: [lʲ e s a] * коли: [k o lʲ i] * слепа: [s lʲ e p 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: 35,137 Examples: * сашки: [s a ʂ kʲ i] * тихую: [tʲ i x u u] * буди: [b u dʲ i] * кипр: [kʲ i p r] |
Occurrences: 8,530 Examples: * зубы: [z u b ɨ] * быть,: [b ɨ tʲ] * коры: [k o r ɨ] * ты…: [t ɨ] |
Occurrences: 17,161 Examples: * пробу: [p r o b u] * ну…: [n u] * тихую: [tʲ i x u u] * зубы: [z u b ɨ] |
||
Close-Mid |
Occurrences: 38,591 Examples: * леща.: [lʲ e ɕː a] * мере.: [mʲ e rʲ e] * кейн.: [kʲ e j n] * мраке: [m r a kʲ e] |
Occurrences: 44,320 Examples: * пробу: [p r o b u] * фонд.: [f o n d] * долгу: [d o l ɡ u] * дочь: [d o t͡ɕʲ] |
|||
Open-Mid |
|||||
Open |
Occurrences: 47,093 Examples: * леща.: [lʲ e ɕː a] * мраке: [m r a kʲ e] * сашки: [s a ʂ kʲ i] * сучья: [s u t͡ɕʲ a] |