Process for improving pronunciation of proper nouns foreign to a target language text-to-speech system

Patent metadata

Field	Value
Publication number	US-2016358596-A1
Application number	US-201514733289-A
Country	US
Kind code	A1
Filing date	Jun 8, 2015
Priority date	Jun 8, 2015
Publication date	Dec 8, 2016
Grant date	—

Abstract

A system and method configured for use in a text-to-speech (TTS) system is provided. Embodiments may include identifying, using one or more processors, a word or phrase as a named entity and identifying a language of origin associated with the named entity. Embodiments may further include transliterating the named entity to a script associated with the language of origin. If the TTS system is operating in the language of origin, embodiments may include passing the transliterated script to the TTS system. If the TTS system is not operating in the language of origin, embodiments may include generating a phoneme sequence in the language of origin using a grapheme to phoneme (G2P) converter.

First claim

What is claimed is:

1 . A computer-implemented method configured for use in a text-to-speech (TTS) system comprising:
identifying, using one or more processors, a word or phrase as a named entity; identifying a language of origin associated with the named entity; transliterating the named entity to a script associated with the language of origin; if the TTS system is operating in the language of origin, passing the transliterated script to the TTS system; and if the TTS system is not operating in the language of origin, generating a phoneme sequence in the language of origin using a grapheme to phoneme (G2P) converter.

2 . The method of claim 1 , further comprising:
if the TTS system is not operating in the language of origin, mapping the phoneme sequence to a sequence of target language phonemes.

3 . The method of claim 2 , wherein mapping includes generating a map of most likely unigram, bigram, and trigram mappings from the phoneme sequence to the sequence of target language phonemes.

4 . The method of claim 1 , wherein identifying a word or phrase as a named entity includes one or more of table lookup and contextual analysis.

5 . The method of claim 1 , wherein identifying a language of origin associated with the named entity includes one or more of table lookup and shortest distance measures to an existing names database.

6 . The method of claim 1 , further comprising:
augmenting a text to speech dictionary based upon, at least in part, the phoneme sequence.

7 . The method of claim 6 , wherein the text to speech dictionary is associated with an automatic speech recognition (ASR) system.

8 . A non-transitory computer-readable storage medium having stored thereon instructions, which when executed by a processor result in one or more operations configured for use in a text-to-speech (TTS) system, the operations comprising:
identifying, using one or more processors, a word or phrase as a named entity; identifying a language of origin associated with the named entity; transliterating the named entity to a script associated with the language of origin; if the TTS system is operating in the language of origin, passing the transliterated script to the TTS system; and if the TTS system is not operating in the language of origin, generating a phoneme sequence in the language of origin using a grapheme to phoneme (G2P) converter.

9 . The non-transitory computer-readable storage medium of claim 8 , further comprising:
if the TTS system is not operating in the language of origin, mapping the phoneme sequence to a sequence of target language phonemes.

10 . The non-transitory computer-readable storage medium of claim 9 , wherein mapping includes generating a map of most likely unigram, bigram, and trigram mappings from the phoneme sequence to the sequence of target language phonemes.

11 . The non-transitory computer-readable storage medium of claim 8 , wherein identifying a word or phrase as a named entity includes one or more of table lookup and contextual analysis.

12 . The non-transitory computer-readable storage medium of claim 8 , wherein identifying a language of origin associated with the named entity includes one or more of table lookup and shortest distance measures to an existing names database.

13 . The non-transitory computer-readable storage medium of claim 8 , further comprising:
augmenting a text to speech dictionary based upon, at least in part, the phoneme sequence.

14 . The non-transitory computer-readable storage medium of claim 13 , wherein the text to speech dictionary is associated with an automatic speech recognition (ASR) system.

15 . A text to speech system comprising:
one or more processors configured to identify a word or phrase as a named entity, the one or more processors further configured to identify a language of origin associated with the named entity and transliterate the named entity to a script associated with the language of origin, if the TTS system is operating in the language of origin, the one or more processors further configured to pass the transliterated script to the TTS system, and if the TTS system is not operating in the language of origin, the one or more processors further configured to generate a phoneme sequence in the language of origin using a grapheme to phoneme (G2P) converter.

16 . The system of claim 15 , wherein if the TTS system is not operating in the language of origin, mapping the phoneme sequence to a sequence of target language phonemes.

17 . The system of claim 16 , wherein mapping includes generating a map of most likely unigram, bigram, and trigram mappings from the phoneme sequence to the sequence of target language phonemes.

18 . The system of claim 15 , wherein identifying a word or phrase as a named entity includes one or more of table lookup and contextual analysis.

19 . The system of claim 15 , wherein identifying a language of origin associated with the named entity includes one or more of table lookup and shortest distance measures to an existing names database.

20 . The system of claim 15 , further comprising:
augmenting a text to speech dictionary based upon, at least in part, the phoneme sequence.

Assignees

• Nuance Communications Inc

Inventors

• Singh Anurag Ratan
• Xu Yifang
• Sanchez Quijas Ivan A
• Godavarti Mahesh

Classifications

• G06F40/295

  Named entity recognition · CPC title

• G10L13/08 Primary

  Text analysis or generation of parameters for speech synthesis out of text, e.g. grapheme to phoneme translation, prosody generation or stress or intonation determination · CPC title

• G06F40/56

  Natural language generation · CPC title

• G06F40/58

  Use of machine translation, e.g. for multi-lingual retrieval, for server-side translation for client devices or for real-time translation · CPC title

• G10L13/086

  Detection of language · CPC title