Ctrl+K
Ctrl+K

Romanian CV dictionary v2.0.0#

  • Maintainer: Vox Communis

  • Language: Romanian

  • Dialect: N/A

  • Phone set: XPF

  • Number of words: 11,388

  • Phones: a b dʒ d̪ e f h i k l̪ m n̪ o p r̪ s s̪ tʃ t̪ t̪s u v z̪ ɜ ɡ ɨ ʃ ʒ

  • License: CC-0

  • Compatible MFA version: v2.0.0

  • Citation:

@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 romanian_cv

Or download from the release page.

Intended use#

This dictionary is intended for forced alignment of Romanian transcripts.

This dictionary uses the XPF phone set for Romanian, and was used in training the Romanian XPF 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

Velar

Glottal

Nasal
Occurrences:
2,862
Examples:
* ampla:
[a m p l̪ a]
* malul:
[m a l̪ u l̪]
* marea:
[m a r̪ e a]
* mint:
[m i n̪ t̪]
Occurrences:
5,638
Examples:
* vino:
[v i n̪ o]
* tren:
[t̪ r̪ e n̪]
* nea:
[n̪ e a]
* mint:
[m i n̪ t̪]

Stop
Occurrences:
2,747
Examples:
* ampla:
[a m p l̪ a]
* rupea:
[r̪ u p e a]
* probe:
[p r̪ o b e]
* rapid:
[r̪ a p i d̪]
Occurrences:
1,074
Examples:
* orbi:
[o r̪ b i]
* bâte:
[b ɨ t̪ e]
* bazac:
[b a z̪ a k]
* probe:
[p r̪ o b e]
Occurrences:
7,193
Examples:
* uite:
[u i t̪ e]
* tren:
[t̪ r̪ e n̪]
* bâte:
[b ɨ t̪ e]
* mint:
[m i n̪ t̪]
Occurrences:
2,058
Examples:
* nodul:
[n̪ o d̪ u l̪]
* de-al:
[d̪ e a l̪]
* rapid:
[r̪ a p i d̪]
* duc:
[d̪ u k]
Occurrences:
3,532
Examples:
* bazac:
[b a z̪ a k]
* clar:
[k l̪ a r̪]
* text:
[t̪ e k s̪ t̪]
* erica:
[e r̪ i k a]
Occurrences:
706
Examples:
* găman:
[ɡ ɜ m a n̪]
* grea:
[ɡ r̪ e a]
* gătit:
[ɡ ɜ t̪ i t̪]
* golul:
[ɡ o l̪ u l̪]

Affricate
Occurrences:
1,025
Examples:
* faci:
[f a tʃ]
* mecea:
[m e tʃ a]
* ucise:
[u tʃ i s̪ e]
* ciclu:
[tʃ i k l̪ u]
Occurrences:
387
Examples:
* logic:
[l̪ o dʒ i k]
* vagi:
[v a dʒ i]
* alege:
[a l̪ e dʒ e]
* lungi:
[l̪ u n̪ dʒ i]

Sibilant
Occurrences:
3,328
Examples:
* să-și:
[s̪ ɜ ʃ i]
* sine:
[s̪ i n̪ e]
* text:
[t̪ e k s̪ t̪]
* simtă:
[s̪ i m t̪ ɜ]
Occurrences:
965
Examples:
* bazac:
[b a z̪ a k]
* zilei:
[z̪ i l̪ e i]
* bazăm:
[b a z̪ ɜ m]
* crez:
[k r̪ e z̪]
Occurrences:
933
Examples:
* ţinem:
[t̪ s i n̪ e m]
* forţa:
[f o r̪ t̪ s a]
* pieţe:
[p i e t̪ s e]
* ţării:
[t̪ s ɜ r̪ i i]
Occurrences:
808
Examples:
* șefii:
[ʃ e f i i]
* să-și:
[s̪ ɜ ʃ i]
* știau:
[ʃ t̪ i a u]
* vișan:
[v i ʃ a n̪]
Occurrences:
256
Examples:
* josh:
[ʒ o s̪ h]
* ajuta:
[a ʒ u t̪ a]
* jucam:
[ʒ u k a m]
* bejan:
[b e ʒ a n̪]

Fricative
Occurrences:
1,037
Examples:
* află:
[a f l̪ ɜ]
* șefii:
[ʃ e f i i]
* faci:
[f a tʃ]
* flori:
[f l̪ o r̪ i]
Occurrences:
1,220
Examples:
* vino:
[v i n̪ o]
* viață:
[v i a t̪s ɜ]
* veni:
[v e n̪ i]
* viță:
[v i t̪s ɜ]
Occurrences:
106
Examples:
* hotel:
[h o t̪ e l̪]
* harry:
[h a r̪ r̪ i]
* josh:
[ʒ o s̪ h]
* haris:
[h a r̪ i s̪]

Lateral
Occurrences:
4,380
Examples:
* ampla:
[a m p l̪ a]
* află:
[a f l̪ ɜ]
* malul:
[m a l̪ u l̪]
* clar:
[k l̪ a r̪]

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:
11,051
Examples:
* orbi:
[o r̪ b i]
* vino:
[v i n̪ o]
* uite:
[u i t̪ e]
* viață:
[v i a t̪s ɜ]
Occurrences:
812
Examples:
* bâte:
[b ɨ t̪ e]
* între:
[ɨ n̪ t̪ r̪ e]
* atâta:
[a t̪ ɨ t̪ a]
* luând:
[l̪ u ɨ n̪ d̪]
Occurrences:
4,264
Examples:
* malul:
[m a l̪ u l̪]
* uite:
[u i t̪ e]
* rupea:
[r̪ u p e a]
* auto:
[a u t̪ o]

Close-Mid
Occurrences:
10,432
Examples:
* uite:
[u i t̪ e]
* tren:
[t̪ r̪ e n̪]
* rupea:
[r̪ u p e a]
* bâte:
[b ɨ t̪ e]
Occurrences:
4,432
Examples:
* orbi:
[o r̪ b i]
* vino:
[v i n̪ o]
* auto:
[a u t̪ o]
* probe:
[p r̪ o b e]

Open-Mid
Occurrences:
2,840
Examples:
* află:
[a f l̪ ɜ]
* viață:
[v i a t̪s ɜ]
* să-și:
[s̪ ɜ ʃ i]
* mină:
[m i n̪ ɜ]

Open
Occurrences:
8,125
Examples:
* ampla:
[a m p l̪ a]
* află:
[a f l̪ ɜ]
* malul:
[m a l̪ u l̪]
* viață:
[v i a t̪s ɜ]
On this page
Show Source