Japanese MFA dictionary v2.0.0#

Maintainer: Montreal Forced Aligner
Language: Japanese
Dialect: N/A
Phone set: MFA
Number of words: 77,695
Phones: a aː b bʲ bː c cː d dz dzː dʑ dʑː dʲ dː e eː h hː i iː i̥ j k kː m mʲ mʲː mː n nː o oː p pʲ pʲː pː s sː t ts tsː tɕ tɕː tʲ tː v vʲ w z ç çː ŋ ɕ ɕː ɟ ɡ ɡː ɨ ɨː ɨ̥ ɯ ɯː ɯ̥ ɰ̃ ɲ ɲː ɴ ɴː ɸ ɸʲ ɸː ɾ ɾʲ ɾː ʑ ʔ
License: CC BY 4.0
Compatible MFA version: v2.0.0
Citation:

@techreport{mfa_japanese_mfa_dictionary_2022,
	author={McAuliffe, Michael and Sonderegger, Morgan},
	title={Japanese MFA dictionary v2.0.0},
	address={\url{https://mfa-models.readthedocs.io/pronunciation dictionary/Japanese/Japanese MFA dictionary v2_0_0.html}},
	year={2022},
	month={Mar},
}

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 japanese_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 Japanese transcripts.

This dictionary uses the MFA phone set for Japanese, and was used in training the Japanese 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	Palatal	Velar	Uvular	Glottal
Nasal	m Occurrences: 17,942 Examples: * すしづめ: [s ɨ̥ ɕ i z ɨ m e] * 蒲公英: [t a m p o p o] * あつめる: [a ts ɨ m e ɾ ɯ] * 見どころ: [m i d o k o ɾ o] mʲ Occurrences: 154 Examples: * ミュンヘン: [mʲ ɨ ɰ̃ h e ɴ] * ミュレット: [mʲ ɨ ɾ e tː o] * みゃくせき: [mʲ a k ɯ s e c i] * 脈あり: [mʲ a k ɯ a ɾ i] mʲː Occurrences: 31 Examples: * ほんみょう: [h o mʲː oː] * ぶんみゃく: [b ɯ mʲː a k ɯ] * 三藐三菩提: [s a mʲː a k ɯ s a m b o d a i] * はんみょう: [h a mʲː oː] mː Occurrences: 540 Examples: * 秋刀魚: [s a mː a] * 勤務先: [c i mː ɯ s a c i] * かんむり座: [k a mː ɯ ɾ i z a] * えんむしつ: [e mː ɯ ɕ i̥ ts ɨ]		n Occurrences: 11,704 Examples: * ネチケット: [n e tɕ i̥ k e tː o] * ウガンダ: [ɯ ɡ a n d a] * 楽天的: [ɾ a k ɯ t e n t e c i] * 伴って: [t o m o n a tː e] nː Occurrences: 380 Examples: * いろんな: [i ɾ o nː a] * 万年筆: [m a nː e ɰ̃ ɸ ɯ d e] * てんのう: [t e nː oː] * 天王星: [t e nː oː s eː]	ɲ Occurrences: 4,985 Examples: * パタゴニア: [p a t a ɡ o ɲ i a] * ざんりゅう: [dz a ɲ ɾʲ ɨː] * インジウム: [i ɲ dʑ i ɯ m ɯ] * くにぐに: [k ɯ ɲ i ɡ ɯ ɲ i] ɲː Occurrences: 155 Examples: * かんにゅう: [k a ɲː ɨː] * ヤンニョム: [j a ɲː o m ɯ] * ばんにん: [b a ɲː i ɴ] * しんにん: [ɕ i ɲː i ɴ]	ŋ Occurrences: 2,993 Examples: * 神経学: [ɕ i ŋ k eː ɡ a k ɯ] * まんけん: [m a ŋ k e ɴ] * とんがる: [t o ŋ ɡ a ɾ ɯ] * 三味線貝: [ɕ a m i s e ŋ ɡ a i]	ɴ Occurrences: 10,160 Examples: * 優柔不断: [j ɨː ʑ ɨː ɸ ɯ d a ɴ] * きずいせん: [c i z ɨ i s e ɴ] * アフガン: [a ɸ ɯ ɡ a ɴ] * さくぶん: [s a k ɯ b ɯ ɴ] ɴː Occurrences: 3 Examples: * うーん: [ɴː]
Stop	p Occurrences: 2,882 Examples: * 蒲公英: [t a m p o p o] * パタゴニア: [p a t a ɡ o ɲ i a] * タイプ: [t a i p ɯ] * テンペラ: [t e m p e ɾ a] pː Occurrences: 715 Examples: * ペッパー: [p e pː aː] * 掻っ払い: [k a pː a ɾ a i] * れっぷ: [ɾ e pː ɯ] * ハラッパー: [h a ɾ a pː aː] b Occurrences: 9,799 Examples: * 海老鯛: [e b i t a i] * こぼす: [k o b o s ɨ] * ベック: [b e kː ɯ] * さくぶん: [s a k ɯ b ɯ ɴ] bː Occurrences: 4 Examples: * ビッベ: [b i bː e] * グッバイ: [ɡ ɯ bː a i] * web会議: [w e bː ɯ k a i ɟ i]		t Occurrences: 16,837 Examples: * 海老鯛: [e b i t a i] * たがやす: [t a ɡ a j a s ɨ] * 蒲公英: [t a m p o p o] * パタゴニア: [p a t a ɡ o ɲ i a] tː Occurrences: 1,366 Examples: * ネチケット: [n e tɕ i̥ k e tː o] * 伴って: [t o m o n a tː e] * さらって: [s a ɾ a tː e] * 上った: [n o b o tː a] d Occurrences: 8,456 Examples: * ウガンダ: [ɯ ɡ a n d a] * 優柔不断: [j ɨː ʑ ɨː ɸ ɯ d a ɴ] * 見どころ: [m i d o k o ɾ o] * ウォード: [w oː d o] dː Occurrences: 99 Examples: * モペッド: [m o p e dː o] * アイポッド: [a i p o dː o] * ミルウッド: [m i ɾ ɯ ɯ dː o] * ゴッデム: [ɡ o dː e m ɯ]	c Occurrences: 9,552 Examples: * 絵描き: [e k a c i] * 支払機: [ɕ i h a ɾ a i c i] * 楽天的: [ɾ a k ɯ t e n t e c i] * きずいせん: [c i z ɨ i s e ɴ] cː Occurrences: 352 Examples: * いっきうち: [i cː i ɯ tɕ i] * はっき: [h a cː i] * 殺菌剤: [s a cː i n dz a i] * 中石器時代: [tɕ ɨː s e cː i ʑ i d a i] ɟ Occurrences: 2,100 Examples: * 御する: [ɟ o s ɨ ɾ ɯ] * ぎゃくすう: [ɟ a k ɯ s ɨː] * きさらぎ: [c i̥ s a ɾ a ɟ i] * かいぎしつ: [k a i ɟ i ɕ i̥ ts ɨ]	k Occurrences: 33,491 Examples: * ネチケット: [n e tɕ i̥ k e tː o] * ldk: [e ɾ ɯ ʑ iː k eː] * 絵描き: [e k a c i] * いくら: [i k ɯ ɾ a] kː Occurrences: 1,584 Examples: * ベック: [b e kː ɯ] * きゃっかん: [c a kː a ɴ] * バッカ: [b a kː a] * しっかり: [ɕ i̥ kː a ɾ i] ɡ Occurrences: 10,172 Examples: * ウガンダ: [ɯ ɡ a n d a] * たがやす: [t a ɡ a j a s ɨ] * パタゴニア: [p a t a ɡ o ɲ i a] * 再来月: [s a ɾ a i ɡ e ts ɨ] ɡː Occurrences: 48 Examples: * すっごい: [s ɨ̥ ɡː o i] * ブルドッグ: [b ɯ ɾ ɯ d o ɡː ɯ] * ツィッギー: [ts i ɡː iː] * big: [b i ɡː ɯ]		ʔ Occurrences: 13 Examples: * いてっ: [i t e ʔ] * カサッ: [k a s a ʔ] * わるっ: [w a ɾ ɯ ʔ] * わあっ: [w aː ʔ]
Affricate			ts Occurrences: 6,666 Examples: * あつめる: [a ts ɨ m e ɾ ɯ] * 対立軸: [t a i ɾ i ts ɨ ʑ i k ɯ] * 再来月: [s a ɾ a i ɡ e ts ɨ] * きぬたこつ: [c i n ɯ t a k o ts ɨ] tsː Occurrences: 77 Examples: * ベッツ: [b a tsː ɨ] * みっつ: [m i tsː ɨ] * くっつく: [k ɯ̥ tsː ɨ k ɯ] * くっつける: [k ɯ̥ tsː ɨ k e ɾ ɯ] dz Occurrences: 1,527 Examples: * ざまみろ: [dz a m a m i ɾ o] * メナンズ: [m e n a n dz ɨ] * ざんりゅう: [dz a ɲ ɾʲ ɨː] * 心臓麻痺: [ɕ i n dz oː m a ç i] dzː Occurrences: 6 Examples: * ドッズ: [d o dzː ɨ] * ヘッズ: [h e dzː ɯ] * グッズ: [ɡ ɯ dzː ɯ] * オッズ: [o dzː ɯ ɴ]	tɕ Occurrences: 5,565 Examples: * ネチケット: [n e tɕ i̥ k e tː o] * ベネチア: [b e n e tɕ i a] * 落ち度: [o tɕ i d o] * 一時停止: [i tɕ i ʑ i t eː ɕ i] tɕː Occurrences: 282 Examples: * マッチ: [m a tɕː i] * っちゃう: [tɕː a a ɯ] * いっち: [i tɕː i] * ニッチ: [ɲ i tɕː i] dʑ Occurrences: 3,176 Examples: * じゅんや: [dʑ ɨ ɰ̃ j a] * インジウム: [i ɲ dʑ i ɯ m ɯ] * ジュール: [dʑ ɨː ɾ ɯ] * もんじん: [m o ɲ dʑ i ɴ] dʑː Occurrences: 34 Examples: * ホッジス: [h o dʑː i s ɨ] * ヘッジス: [h e dʑː i s ɨ] * ブリッジ: [b ɯ ɾ i dʑː i] * ハッジ: [h a dʑː i]
Sibilant			s Occurrences: 18,766 Examples: * すしづめ: [s ɨ̥ ɕ i z ɨ m e] * たがやす: [t a ɡ a j a s ɨ] * せいたい: [s eː t a i] * こぼす: [k o b o s ɨ] sː Occurrences: 430 Examples: * ほっする: [h o sː ɨ ɾ ɯ] * じっせん: [dʑ i sː e ɴ] * 血清学: [k e sː eː ɡ a k ɯ] * いっすん: [i sː ɨ ɴ] z Occurrences: 3,604 Examples: * すしづめ: [s ɨ̥ ɕ i z ɨ m e] * きずいせん: [c i z ɨ i s e ɴ] * 脊髄反射: [s e c i z ɨ i h a ɰ̃ ɕ a] * ふしぜん: [ɸ ɯ ɕ i z e ɴ]	ɕ Occurrences: 14,528 Examples: * すしづめ: [s ɨ̥ ɕ i z ɨ m e] * 支払機: [ɕ i h a ɾ a i c i] * 観測者: [k a ɰ̃ s o k ɯ ɕ a] * 衝動買い: [ɕ oː d oː ɡ a i] ɕː Occurrences: 447 Examples: * れっし: [ɾ e ɕː i] * ぎっしゃ: [ɟ i ɕː a] * 夜行列車: [j a k oː ɾ e ɕː a] * まっしょう: [m a ɕː oː] ʑ Occurrences: 3,858 Examples: * ldk: [e ɾ ɯ ʑ iː k eː] * 対立軸: [t a i ɾ i ts ɨ ʑ i k ɯ] * 優柔不断: [j ɨː ʑ ɨː ɸ ɯ d a ɴ] * くじゅう: [k ɯ ʑ ɨː]
Fricative	ɸ Occurrences: 3,342 Examples: * 優柔不断: [j ɨː ʑ ɨː ɸ ɯ d a ɴ] * フレスノ: [ɸ ɯ ɾ e s ɨ n o] * アフガン: [a ɸ ɯ ɡ a ɴ] * フクシア: [ɸ ɯ̥ k ɯ ɕ i a] ɸʲ Occurrences: 7 Examples: * フュール: [ɸʲ ɨː ɾ ɯ] * フューリー: [ɸʲ ɨː ɾ iː] ɸː Occurrences: 18 Examples: * ハッフル: [h a ɸː ɯ ɾ ɯ] * スタッフ: [s ɨ̥ t a ɸː ɯ] * シャッフル: [ɕ a ɸː ɯ ɾ ɯ] * イッフィー: [i ɸː iː]	v Occurrences: 198 Examples: * ヴァレーゼ: [v a ɾ eː z e] * マハデヴィ: [m a h a d e v i] * ヴィシュヌ: [v i ɕ ɨ n ɯ] * マクヴィ: [m a k ɯ v i] vʲ Occurrences: 4 Examples:		ç Occurrences: 2,587 Examples: * 心臓麻痺: [ɕ i n dz oː m a ç i] * 飛行士: [ç i̥ k oː ɕ i] * ショーヒン: [ɕ oː ç i ɴ] * 手のひら: [t e n o ç i ɾ a] çː Occurrences: 3 Examples:			h Occurrences: 6,282 Examples: * 支払機: [ɕ i h a ɾ a i c i] * 脊髄反射: [s e c i z ɨ i h a ɰ̃ ɕ a] * はしごしゃ: [h a ɕ i ɡ o ɕ a] * くみはん: [k ɯ m i h a ɴ] hː Occurrences: 5 Examples: * ゼンパッハ: [dz e m p a hː a] * ウンバッハ: [ɯ m b a hː a] * ゴッホ: [ɡ o hː o] * ゲルラッハ: [ɡ e ɾ ɯ ɾ a hː a]
Approximant	w Occurrences: 3,155 Examples: * ウォード: [w oː d o] * 訳あり: [w a k e a ɾ i] * ためらわ: [t a m e ɾ a w a] * 我が家: [w a ɡ a j a]			j Occurrences: 6,118 Examples: * たがやす: [t a ɡ a j a s ɨ] * 優柔不断: [j ɨː ʑ ɨː ɸ ɯ d a ɴ] * やまいだれ: [j a m a i d a ɾ e] * ぬるま湯: [n ɯ ɾ ɯ m a j ɨ]	ɰ̃ Occurrences: 3,962 Examples: * 観測者: [k a ɰ̃ s o k ɯ ɕ a] * 脊髄反射: [s e c i z ɨ i h a ɰ̃ ɕ a] * フェンス: [ɸ e ɰ̃ s ɨ] * じゅんや: [dʑ ɨ ɰ̃ j a]
Tap			ɾ Occurrences: 24,817 Examples: * ldk: [e ɾ ɯ ʑ iː k eː] * いくら: [i k ɯ ɾ a] * あつめる: [a ts ɨ m e ɾ ɯ] * 対立軸: [t a i ɾ i ts ɨ ʑ i k ɯ] ɾʲ Occurrences: 1,264 Examples: * ざんりゅう: [dz a ɲ ɾʲ ɨː] * りゅうおん: [ɾʲ ɨː o ɴ] * 荒療治: [a ɾ a ɾʲ oː ʑ i] * 円電流: [e n d e ɲ ɾʲ ɨː] ɾː Occurrences: 4 Examples: * ビツェッリ: [b i ts e ɾː i] * バガレッラ: [b a ɡ a ɾ e ɾː a] * チッリ: [tɕ i ɾː i]

Vowels#

Vowel symbols to the left of are unrounded and those to the right are rounded.

	Front	Central	Back
Close	i Occurrences: 52,344 Examples: * 海老鯛: [e b i t a i] * すしづめ: [s ɨ̥ ɕ i z ɨ m e] * 絵描き: [e k a c i] * いくら: [i k ɯ ɾ a] iː Occurrences: 2,512 Examples: * ldk: [e ɾ ɯ ʑ iː k eː] * ときいろ: [t o c iː ɾ o] * いちいち: [i tɕ iː tɕ i] * 差し入れ: [s a ɕ iː ɾ e] i̥ Occurrences: 3,261 Examples: * ネチケット: [n e tɕ i̥ k e tː o] * きさらぎ: [c i̥ s a ɾ a ɟ i] * 位相幾何学: [i s oː c i̥ k a ɡ a k ɯ] * かいぎしつ: [k a i ɟ i ɕ i̥ ts ɨ]	ɨ Occurrences: 14,309 Examples: * すしづめ: [s ɨ̥ ɕ i z ɨ m e] * たがやす: [t a ɡ a j a s ɨ] * あつめる: [a ts ɨ m e ɾ ɯ] * 対立軸: [t a i ɾ i ts ɨ ʑ i k ɯ] ɨː Occurrences: 4,540 Examples: * 優柔不断: [j ɨː ʑ ɨː ɸ ɯ d a ɴ] * くじゅう: [k ɯ ʑ ɨː] * ざんりゅう: [dz a ɲ ɾʲ ɨː] * ぎゃくすう: [ɟ a k ɯ s ɨː] ɨ̥ Occurrences: 2,906 Examples: * すしづめ: [s ɨ̥ ɕ i z ɨ m e] * すっごい: [s ɨ̥ ɡː o i] * ステラー: [s ɨ̥ t e ɾ aː] * 捨て仮名: [s ɨ̥ t e ɡ a n a]	ɯ Occurrences: 30,079 Examples: * ウガンダ: [ɯ ɡ a n d a] * ldk: [e ɾ ɯ ʑ iː k eː] * いくら: [i k ɯ ɾ a] * あつめる: [a ts ɨ m e ɾ ɯ] ɯː Occurrences: 729 Examples: * ルール: [ɾ ɯː ɾ ɯ] * クーパー: [k ɯː p aː] * たいぐう: [t a i ɡ ɯː] * ナブーム: [n a b ɯː m ɯ] ɯ̥ Occurrences: 1,989 Examples: * フクシア: [ɸ ɯ̥ k ɯ ɕ i a] * 説き伏せる: [t o c i ɸ ɯ̥ s e ɾ ɯ] * ふくつ: [ɸ ɯ̥ k ɯ ts ɨ] * やくしゃ: [j a k ɯ̥ ɕ a]

Close-Mid	e Occurrences: 25,791 Examples: * ネチケット: [n e tɕ i̥ k e tː o] * ldk: [e ɾ ɯ ʑ iː k eː] * 海老鯛: [e b i t a i] * すしづめ: [s ɨ̥ ɕ i z ɨ m e] eː Occurrences: 6,428 Examples: * ldk: [e ɾ ɯ ʑ iː k eː] * せいたい: [s eː t a i] * キャンデー: [c a n d eː] * 神経学: [ɕ i ŋ k eː ɡ a k ɯ]		o Occurrences: 34,252 Examples: * ネチケット: [n e tɕ i̥ k e tː o] * 蒲公英: [t a m p o p o] * パタゴニア: [p a t a ɡ o ɲ i a] * こぼす: [k o b o s ɨ] oː Occurrences: 17,658 Examples: * ウォード: [w oː d o] * 衝動買い: [ɕ oː d oː ɡ a i] * 運動会: [ɯ n d oː k a i] * とおり: [t oː ɾ i]

Open-Mid

Open		a Occurrences: 64,942 Examples: * ウガンダ: [ɯ ɡ a n d a] * 海老鯛: [e b i t a i] * たがやす: [t a ɡ a j a s ɨ] * 絵描き: [e k a c i] aː Occurrences: 2,382 Examples: * ステラー: [s ɨ̥ t e ɾ aː] * ハイパー: [h a i p aː] * ステッカー: [s ɨ̥ t e kː aː] * ペッパー: [p e pː aː]