Hindi CV dictionary v2.0.0#

Maintainer: Vox Communis
Language: Hindi
Dialect: N/A
Phone set: Epitran
Number of words: 11,113
Phones: a aː b b̤ d d̤ d͡ʒ d͡ʒ̤ e f h i iː j k kʰ l m n o p pʰ q r r̩ s t tʰ t͡ʃ t͡ʃʰ u uː v w x z ã ãː æ æː ŋ ũ ũː ɑ ɔ ɔː ɖ ɖ̤ ə ə̃ ɛ ɡ ɡ̤ ɣ ɦ ɪ ɲ ɳ ɹ ɹ̩ ɽ ɽ̥ ʂ ʃ ʈ ʈʰ ʊ ʌ θ ẽ
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}
}

If you have comments or questions about this dictionary or its phone set, you can check previous MFA model discussion posts or create a new one.

Installation#

Install from the MFA command line:

mfa model download dictionary hindi_cv

Or download from the release page.

Intended use#

This dictionary is intended for forced alignment of Hindi transcripts.

This dictionary uses the Epitran phone set for Hindi, and was used in training the Hindi 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	Dental	Alveolar	Alveopalatal	Retroflex	Palatal	Velar	Uvular	Glottal
Nasal	m Occurrences: 2,128 Examples: * मजा,: [m ə d͡ʒ aː] * 'लंबी: [l ə m b iː] * कॉम: [k ɔ m] * समय: [s ə m ə j]			n Occurrences: 4,001 Examples: * फांस: [pʰ aː n s ə] * पूछने: [p uː t͡ʃʰ n e] * शांति: [ʃ aː n t i] * अनिल,: [a n i l]		ɳ Occurrences: 175 Examples: * रहाणे: [r ə ɦ aː ɳ e] * घृणित: [ɡ̤ r̩ ɳ i t] * जाणा: [d͡ʒ aː ɳ aː] * जागरण: [d͡ʒ aː ɡ r ə ɳ]	ɲ Occurrences: 109 Examples: * मंच: [m ə ɲ t͡ʃ ə] * मांझी: [m aː ɲ d͡ʒ̤ iː] * भांजे: [b̤ aː ɲ d͡ʒ e] * ऊंचा: [uː ɲ t͡ʃ aː]	ŋ Occurrences: 463 Examples: * आएंगी: [aː e ŋ ɡ iː] * बोइंग: [b o i ŋ ɡ ə] * सिंगल: [s i ŋ ɡ ə l] * संघ: [s ə ŋ ɡ̤ ə]
Stop Plain	p Occurrences: 1,526 Examples: * पूछने: [p uː t͡ʃʰ n e] * पैरोल: [p æː r o l] * पिता: [p i t aː] * टिप्स: [ʈ i p s ə] b Occurrences: 1,348 Examples: * बाज़ी: [b aː z iː] * टेबल,: [ʈ e b ə l] * बीएड: [b iː e ɖ] * सुबोध: [s u b o d̤]			t Occurrences: 1,900 Examples: * शांति: [ʃ aː n t i] * जीते।: [d͡ʒ iː t e] * तबीयत: [t ə b iː j ə t] * पिता: [p i t aː] d Occurrences: 1,293 Examples: * देखें: [d e kʰ e n] * दिन: [d i n] * खाद्य: [kʰ aː d j ə] * दौलत: [d ɔː l ə t]		ʈ Occurrences: 1,140 Examples: * टेबल,: [ʈ e b ə l] * टिप्स: [ʈ i p s ə] * होस्ट: [ɦ o s ʈ ə] * माटी: [m aː ʈ iː] ɖ Occurrences: 606 Examples: * बीएड: [b iː e ɖ] * सिडनी: [s i ɖ n iː] * गोल्ड: [ɡ o l ɖ ə] * कुंडू: [k u n ɖ uː]		k Occurrences: 2,480 Examples: * ग्रीक: [ɡ r iː k] * किया:: [k i j aː] * रोका।: [r o k aː] * साईकल: [s aː iː k ə l] ɡ Occurrences: 1,356 Examples: * ग्रीक: [ɡ r iː k] * आएंगी: [aː e ŋ ɡ iː] * बोइंग: [b o i ŋ ɡ ə] * रखेगा: [r ə kʰ e ɡ aː]	q Occurrences: 16 Examples: * ताक़त: [t aː q ə t] * क़ैदी: [q æː d iː] * वक़्त: [v ə q t ə] * क़त्ल: [q ə t l ə]
Aspirated	b̤ Occurrences: 324 Examples: * 'भाबी: [b̤ aː b iː] * भट्टी: [b̤ ə ʈ ʈ iː] * भड़क: [b̤ ə ɽ ə k] * निभा: [n i b̤ aː]			d̤ Occurrences: 317 Examples: * सुबोध: [s u b o d̤] * कंधा: [k ə n d̤ aː] * धूम: [d̤ uː m] * बर्धन: [b ə r d̤ ə n]		ɖ̤ Occurrences: 28 Examples: * ढोल: [ɖ̤ o l] * गड्ढा: [ɡ ə ɖ ɖ̤ aː] * बढावा: [b ə ɖ̤ aː v aː] * गड्ढे: [ɡ ə ɖ ɖ̤ e]		ɡ̤ Occurrences: 94 Examples: * घोसले: [ɡ̤ o s l e] * संघ: [s ə ŋ ɡ̤ ə] * घृणित: [ɡ̤ r̩ ɳ i t] * घटाई: [ɡ̤ ə ʈ aː iː]
Affricate Plain					t͡ʃ Occurrences: 686 Examples: * मंच: [m ə ɲ t͡ʃ ə] * चीन,: [t͡ʃ iː n] * चोवा: [t͡ʃ o v aː] * चुनो।: [t͡ʃ u n o] d͡ʒ Occurrences: 1,197 Examples: * जीते।: [d͡ʒ iː t e] * सीरीज: [s iː r iː d͡ʒ] * मजा,: [m ə d͡ʒ aː] * जॉन: [d͡ʒ ɔ n]
Aspirated					d͡ʒ̤ Occurrences: 90 Examples: * मुझपर: [m u d͡ʒ̤ p ə r] * मुरझा: [m u r d͡ʒ̤ aː] * मांझी: [m aː ɲ d͡ʒ̤ iː] * साझा: [s aː d͡ʒ̤ aː]
Sibilant				s Occurrences: 2,396 Examples: * फांस: [pʰ aː n s ə] * सिर्फ: [s i r pʰ ə] * सीरीज: [s iː r iː d͡ʒ] * सुबोध: [s u b o d̤] z Occurrences: 121 Examples: * बाज़ी: [b aː z iː] * ज़रूर: [z ə r uː r] * आवाज़: [aː v aː z] * नाज़ी: [n aː z iː]	ʃ Occurrences: 705 Examples: * शांति: [ʃ aː n t i] * शरमा: [ʃ ə r m aː] * शक्कर: [ʃ ə k k ə r] * शीना: [ʃ iː n aː]	ʂ Occurrences: 202 Examples: * पक्षी: [p ə k ʂ iː] * परिषद: [p ə r i ʂ ə d] * विशेष: [v i ʃ e ʂ] * मनीष: [m ə n iː ʂ]
Fricative		f Occurrences: 88 Examples: * तरफ़: [t ə r ə f] * photo: [f o w t o w] * first: [f ɹ̩ s t] * फ़ूल: [f uː l] v Occurrences: 1,246 Examples: * विरार: [v i r aː r] * पर्व: [p ə r v ə] * विन: [v i n] * सवाल:: [s ə v aː l]	θ Occurrences: 1 Examples:							h Occurrences: 102 Examples: * लगाईः: [l ə ɡ aː iː ə h] * कपः: [k ə p ə h] * केरलः: [k e r l ə h] * रूमः: [r uː m ə h] ɦ Occurrences: 1,163 Examples: * महफूज: [m ə ɦ pʰ uː d͡ʒ] * हंस: [ɦ ə n s ə] * नही: [n ə ɦ iː] * हाइप: [ɦ aː i p]
Approximant	w Occurrences: 30 Examples: * photo: [f o w t o w] * world: [w ɹ̩ l d] * rayo: [ɹ e j o w] * zero: [z ɪ ɹ o w]			ɹ Occurrences: 34 Examples: * rayo: [ɹ e j o w] * redmi: [ɹ ɛ d m i] * heart: [h ɑ ɹ t] * wrap:: [ɹ æ p] ɹ̩ Occurrences: 12 Examples: * first: [f ɹ̩ s t] * world: [w ɹ̩ l d]			j Occurrences: 1,345 Examples: * लिये: [l i j e] * किया:: [k i j aː] * तबीयत: [t ə b iː j ə t] * समय: [s ə m ə j]
Tap						ɽ Occurrences: 299 Examples: * पकड़ा: [p ə k ɽ aː] * पकड़े: [p ə k ɽ e] * भड़क: [b̤ ə ɽ ə k] * बछड़े: [b ə t͡ʃʰ ɽ e] ɽ̥ Occurrences: 76 Examples: * पढ़के: [p ə ɽ̥ ə k e] * पढ़ाई: [p ə ɽ̥ aː iː] * पढ़ता: [p ə ɽ̥ ə t aː] * पढ़ती: [p ə ɽ̥ ə t iː]
Trill				r Occurrences: 4,058 Examples: * सिर्फ: [s i r pʰ ə] * विरार: [v i r aː r] * ग्रीक: [ɡ r iː k] * सीरीज: [s iː r iː d͡ʒ] r̩ Occurrences: 51 Examples: * घृणित: [ɡ̤ r̩ ɳ i t] * मातृ: [m aː t r̩] * वृद्ध: [v r̩ d d̤ ə] * मृत: [m r̩ t]
Lateral				l Occurrences: 2,313 Examples: * अनिल,: [a n i l] * लिये: [l i j e] * टेबल,: [ʈ e b ə l] * लिखी?: [l i kʰ iː]

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	i Occurrences: 2,711 Examples: * सिर्फ: [s i r pʰ ə] * शांति: [ʃ aː n t i] * विरार: [v i r aː r] * अनिल,: [a n i l] iː Occurrences: 2,816 Examples: * जीते।: [d͡ʒ iː t e] * बाज़ी: [b aː z iː] * ग्रीक: [ɡ r iː k] * लिखी?: [l i kʰ iː]				u Occurrences: 1,262 Examples: * सुबोध: [s u b o d̤] * कुंडू: [k u n ɖ uː] * बुनना: [b u n n aː] * कुरसी: [k u r s iː] uː Occurrences: 632 Examples: * पूछने: [p uː t͡ʃʰ n e] * ज़रूर: [z ə r uː r] * महफूज: [m ə ɦ pʰ uː d͡ʒ] * रूप: [r uː p]
		ɪ Occurrences: 39 Examples: * india: [ɪ n d i ə] * bigg: [b ɪ ɡ] * ltd: [l ɪ m ɪ t ɪ d] * zero: [z ɪ ɹ o w]		ʊ Occurrences: 3 Examples: * look:: [l ʊ k]
Close-Mid	e Occurrences: 2,940 Examples: * पूछने: [p uː t͡ʃʰ n e] * जीते।: [d͡ʒ iː t e] * लिये: [l i j e] * टेबल,: [ʈ e b ə l]				o Occurrences: 1,631 Examples: * रोका।: [r o k aː] * सुबोध: [s u b o d̤] * पैरोल: [p æː r o l] * रोनी: [r o n iː]
			ə Occurrences: 6,777 Examples: * फांस: [pʰ aː n s ə] * सिर्फ: [s i r pʰ ə] * टेबल,: [ʈ e b ə l] * ज़रूर: [z ə r uː r]
Open-Mid	ɛ Occurrences: 19 Examples: * get: [ɡ ɛ t] * tech: [t ɛ k] * celeb: [s ɛ l ɛ b] * redmi: [ɹ ɛ d m i]				ʌ Occurrences: 8 Examples: * what: [w ʌ t] * love: [l ʌ v] ɔ Occurrences: 218 Examples: * जॉन: [d͡ʒ ɔ n] * कॉम: [k ɔ m] * ब्रॉड: [b r ɔ ɖ] * लॉकर: [l ɔ k ə r] ɔː Occurrences: 177 Examples: * सौ,: [s ɔː] * दौलत: [d ɔː l ə t] * बौद्ध: [b ɔː d d̤ ə] * रौनक: [r ɔː n ə k]
	æ Occurrences: 23 Examples: * apple: [æ p ə l] * wrap:: [ɹ æ p] * pak: [p æ k] * app: [æ p] æː Occurrences: 430 Examples: * पैरोल: [p æː r o l] * कैन: [k æː n] * बैठे: [b æː ʈʰ e] * फैन्स: [pʰ æː n s ə]
Open			a Occurrences: 402 Examples: * अनिल,: [a n i l] * अनवर: [a n v ə r] * अदा: [a d aː] * अवश्य: [a v ʃ j ə] aː Occurrences: 6,338 Examples: * फांस: [pʰ aː n s ə] * शांति: [ʃ aː n t i] * विरार: [v i r aː r] * बाज़ी: [b aː z iː]		ɑ Occurrences: 35 Examples: * pok: [p ɑ k] * khan: [k ɑ n] * heart: [h ɑ ɹ t] * stock: [s t ɑ k]

Nasal Vowels#

	Front	Central	Back
Close			ũ Occurrences: 15 Examples: * मुँह: [m ũ ɦ ə] * पहुँच: [p ə ɦ ũ t͡ʃ ə] * उँगली: [ũ ɡ ə l iː] ũː Occurrences: 56 Examples: * आऊँगा: [aː ũː ɡ aː] * हूँ": [ɦ ũː] * जाऊँ।: [d͡ʒ aː ũː] * ऊँची: [ũː t͡ʃ iː]

Close-Mid	ẽ Occurrences: 14 Examples: * आएँ।: [aː ẽ] * आएँगे: [aː ẽ ɡ e] * लाएँ: [l aː ẽ]
		ə̃ Occurrences: 4 Examples: * फँस: [pʰ ə̃ s ə] * हँसे।: [ɦ ə̃ s e] * जँचता: [d͡ʒ ə̃ t͡ʃ ə t aː]
Open-Mid

Open		ã Occurrences: 3 Examples: * अँडे: [ã ɖ e] ãː Occurrences: 72 Examples: * माँग: [m ãː ɡ ə] * आँखे: [ãː kʰ e] * छलाँग: [t͡ʃʰ ə l ãː ɡ ə] * टाँगा: [ʈ ãː ɡ aː]