Method and device for phonetizing data sets containing text

• US 9,436,675 B2
• Filed: 02/11/2013
• Issued: 09/06/2016
• Est. Priority Date: 02/16/2012
• Status: Active Grant

• Current Assignee: Continental Automotive GmbH (Continental AG)
• Original Assignee: Continental Automotive GmbH (Continental AG)
• Inventors: Walther, Jens
• Primary Examiner(s): Desir, Pierre-Louis
• Assistant Examiner(s): THOMAS-HOMESCU, ANNE L
• Application Number:
  US14/378,896
• Publication Number:
  US 20150302001A1
• Time in Patent Office: 1,303 Days
• Field of Search: 704/252, 704/209, 704/260, 704/254, 704/275, 704/211, 704/251, 704/236, 704/2, 704/258, 704/235, 704/243, 704/10, 704/220, 704/266, 704/231, 704/257
• US Class Current: 1/1
• CPC Class Codes: G06F 40/30 Semantic analysis
• CPC Class Codes: G10L 13/04 Details of speech synthesis systems, e.g. synthesiser structure or memory management
• CPC Class Codes: G10L 13/08 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 Class Codes: G10L 15/08 Speech classification or search
• CPC Class Codes: G10L 15/18 using natural language modelling
• CPC Class Codes: G10L 15/187 Phonemic context, e.g. pronunciation rules, phonotactical constraints or phoneme n-grams
• CPC Class Codes: G10L 15/22 Procedures used during a speech recognition process, e.g. man-machine dialogue

Abstract

A method for phonetizing text-containing data records that include graphemes includes: phonetizing the data records by converting the graphemes in the data records into phonemes, and storing the phonemes as phonetized data records; and preprocessing to condition the graphemes for the phonetization by modifying the graphemes on a language-defined and/or user-defined basis. The preprocessing of the graphemes and the conversion of the graphemes into phonemes are performed in parallel on different computation units or different portions of the computation units.

Claims

1. A method for phonetizing text-containing data records (2) that include graphemes, and which are stored in a first storage, by a voice-controlled user interface at a vehicle having a computation device (4) having (a) a first computation unit (5), and (b) a second computation unit (6), the first and second computation units being configured to operate in parallel with one another, the second computation unit (6) having a voice recognition unit, and wherein the data records comprise music information, the method comprising:

the voice-controlled user interface generating phonetized data lists (8) by converting graphemes in the data records (2) stored in the first storage into phonemes, and storing, in a second storage, the phonemes as the phonetized data lists (8), the generating comprising;

breaking down the data records (2) present as graphemes into grapheme subpackets, wherein a length of a grapheme subpacket is determined by an optimum ratio dependent on the first computation unit (5), for a volume of data in the grapheme subpacket to a messaging overhead that arises from the communication between the first and second computation units (5, 6), wherein the message overhead represents the computational power of the first and second computation units based on the grapheme subpackets that are interchanged between the first and the second computation units;

preprocessing, by the first computation unit (5), the graphemes to condition the graphemes for the phonetization by modifying the grapheme subpackets on a language-defined and/or user-defined basis,

converting, by the second computation unit (6), the preprocessed grapheme subpackets into phonemes, wherein the preprocessing of the grapheme subpackets and the conversion of the grapheme subpackets into phonemes are performed in parallel on the first and second computation units (5, 6);

receiving user speech for identifying a selection of music;

performing, by the voice recognition unit, voice recognition on the received speech based on the converting of the preprocessed grapheme subpackets into phonemes, wherein the voice recognition unit is located at the second computation unit; and

playing back, at the vehicle, a music track based on the recognized speech and the converted grapheme subpackets.

2. The method as claimed in claim 1, wherein the preprocessing comprises grammar-based parsing.

3. The method as claimed in claim 1, wherein the preprocessing uses an acoustic model for grapheme-to-phoneme conversion, the method further comprising converting characters not supported by the acoustic model into grapheme symbols supported by the acoustic model.

4. A device for phonetizing text-containing data records (2) that include graphemes by carrying out the method of claim 1, the device comprising:

a data interface (3) configured to input the text-containing data records (2); and

the computation device (4) configured to convert the graphemes in the data records into phonemes and to preprocess the graphemes,

wherein the computation device (4) has at least the first computation unit (5) and the second computation unit (6), the first and second computation units (5, 6).

5. A non-transitory computer readable medium storing program code that, when executed, controls a computation device (4) of a device (1) for phonetizing text-containing data records (2) having at least first and second computation units (5, 6) to carry out the method as claimed in claim 1.