Voice Recognition Dictionary Generation Apparatus and Voice Recognition Dictionary Generation Method

• US 20100299143A1
• Filed: 02/18/2010
• Published: 11/25/2010
• Est. Priority Date: 05/22/2009
• Status: Active Grant

• Current Assignee: Alpine Electronics Incorporated (Alps Alpine Co., Ltd.)
• Original Assignee: Alpine Electronics Incorporated (Alps Alpine Co., Ltd.)
• Inventors: Aoyama, Fumihiko, Takeda, Chiharu
• Granted Patent: US 8,706,484 B2
• Time in Patent Office: Days
• US Class Current: 704/10
• CPC Class Codes: G10L 15/065 Adaptation
• CPC Class Codes: G10L 2015/0635 updating or merging of old and new templates; Mean values; Weighting

Abstract

A voice recognition dictionary generation apparatus and method for suppressing reduction of processing speed at the time of updating. The apparatus includes an input unit configured to receive a text subjected to voice recognition, a storage unit configured to store the text with respect to each file of a predetermined item, a reading data generation unit configured to analyze the text and generate a reading data, and a voice recognition dictionary configured to include content dictionaries that store therein the reading data of the text with respect to each file of the predetermined item. When the file of the predetermined item including the text stored in the storage unit is updated, a control unit detects a total number of the content dictionaries, and when the total number is smaller than a predetermined limit, the control unit generates the content dictionaries with respect to each updated predetermined item.

Claims

1. A voice recognition dictionary generation apparatus comprising:

an input unit configured to receive a text subjected to voice recognition from an external device or a medium;

a storage unit configured to store therein the text with respect to each file of a predetermined item;

a reading data generation unit configured to analyze the text and generate a reading data associated with a pronunciation of the text;

a voice recognition dictionary configured to include content dictionaries, wherein the reading data associated with the text is stored in the content dictionaries with respect to each file of the predetermined item; and

a control unit configured to store the text received via the input unit in the storage unit, wherein

when a file of the predetermined item including the text stored in the storage unit is updated, the control unit detects a total number of the content dictionaries, and when the total number of the content dictionaries is smaller than a predetermined limit, the control unit generates the content dictionaries with respect to each updated predetermined item, thereby updating the voice recognition dictionary.

2. A voice recognition dictionary generation apparatus according to claim 1, wherein, when the total number of the content dictionaries is equal to or larger than a predetermined limit, the control unit deletes the content dictionaries corresponding to the external device or the medium from the voice recognition dictionary, extracts all texts subjected to the voice recognition corresponding the external device or the medium from the storage unit with respect to each content, and generates the content dictionaries, thereby updating the voice recognition dictionary.

3. A voice recognition dictionary generation apparatus according to claim 2, wherein the control unit integrates texts with the same name into a single text among the text subjected to the voice recognition to generate the content dictionaries.

4. A voice recognition dictionary generation apparatus according to claim 1, wherein, when the external device or the medium is connected and the text is stored in the file of the predetermined item in the storage unit, the control unit determines that the file of the predetermined item is updated.

5. A voice recognition dictionary generation apparatus according to claim 4, wherein, when the medium is a compact disc (CD), the predetermined item is music information including a title, an artist name, an album title, a composer name, and a genre.

6. A voice recognition dictionary generation apparatus according to claim 4, wherein the external device is a digital audio device or a mobile phone, and the medium is a CD or a digital versatile disc (DVD).

7. A voice recognition dictionary generation apparatus according to claim 6, wherein, in the case where the medium is a CD, the control unit determines that information is updated when there is no artist name that is matched with the music information stored in a item of the artist name in the storage unit, or when there is no composer name that is matched with the music information stored in the item of the composer name in the storage unit.

8. A voice recognition dictionary generation apparatus according to claim 6, wherein, in the case where the medium is a CD, the control unit determines that information is updated when there is no album title that is matched with the music information stored in a item of the album name in the storage unit, or when TOC information is different from that of a matching album title.

9. A voice recognition dictionary generation method comprising the steps of:

determining whether information in a file of a predetermined item including a text subjected to voice recognition stored in a storage unit is updated;

detecting a total number of content dictionaries that store therein reading data of the text with respect to each file of the predetermined item when the information in the file of the predetermined item including the text obtained from an external device or a medium and subjected to the voice recognition is updated;

comparing the total number of the content dictionaries with a predetermined limit;

extracting the text subjected to the voice recognition corresponding to the external device or the medium with respect to each predetermined item that is updated from the storage unit when the total number of the content dictionaries is smaller than the predetermined limit;

obtaining the reading data by converting the text; and

updating a voice recognition dictionary by storing the reading data obtained by converting the text.

10. A voice recognition dictionary generation method according to claim 9, after comparing the total number of the content dictionaries with a predetermined limit, further comprising:

deleting the content dictionary corresponding to the external device or the medium from the voice recognition dictionary when the total number of the content dictionaries is equal to or larger than the predetermined limit;

extracting all texts subjected to the voice recognition corresponding to the external device or the medium from the storage unit with respect to each predetermined item;

obtaining the reading data by converting the texts; and

updating a voice recognition dictionary by storing the reading data obtained by converting the texts.

11. A voice recognition dictionary generation method according to claim 10, wherein obtaining the reading data includes:

detecting texts having an identical name from among the texts subjected to the voice recognition; and

integrating the texts having the identical name into a single text.

12. A voice recognition dictionary generation method according to claim 9, wherein determining whether the information in the file of the predetermined item is updated is performed by determining whether the external device or the medium is connected and whether the text is stored in the file of the predetermined item in the storage unit.

13. A voice recognition dictionary generation method according to claim 12, wherein, when the medium is a CD, the predetermined item is music information including a title, an artist name, an album title, a composer name, and a genre.

14. A voice recognition dictionary generation method according to claim 12, wherein the external device is a digital audio device or a mobile phone, and the medium is a CD or a DVD.

15. A voice recognition dictionary generation method according to claim 14, wherein, in the case where the medium is a CD, determining whether the information in the file of the predetermined item is updated includes:

extracting an artist name from the predetermined item;

determining whether the artist name is stored in the storage unit; and

determining that the information is updated when there is no identical artist name.

16. A voice recognition dictionary generation method according to claim 14, wherein, in the case where the medium is CD, determining whether the information in the file of the predetermined item is updated includes:

extracting a composer name from the predetermined item;

determining whether the composer name is stored in the storage unit; and

determining that the information is updated when there is no identical composer name.

17. A voice recognition dictionary generation method according to claim 14, wherein, in the case where the medium is a CD, determining whether the information in the file of the predetermined item is updated includes:

extracting an album title from the predetermined item;

determining whether the album title is stored in the storage unit; and

determining that the information is updated when there is no identical album title.

18. A voice recognition dictionary generation method according to claim 14, wherein, in the case where the medium is a CD, determining whether the information in the file of the predetermined item is updated includes:

extracting an album title from the predetermined item; and

determining that the information is updated when TOC information is different from that of a matching album title stored in the storage unit.

19. A voice recognition dictionary generation apparatus comprising:

a storage unit configured to store therein text subjected to voice recognition with respect to each file of a predetermined item, the text being received from an external device;

a data generation unit configured to analyze the text and generate pronunciation data associated with the text;

a voice recognition dictionary configured to include content dictionaries, wherein the pronunciation data associated with the text is stored in the content dictionaries with respect to each file of the predetermined item; and

a control unit configured to store the text received from the external device in the storage unit, wherein

when a file of the predetermined item including the text stored in the storage unit is updated, the control unit detects a total number of the content dictionaries, and when the total number of the content dictionaries is smaller than a predetermined limit, the control unit generates new content dictionaries with respect to each updated predetermined item and stores the new content dictionaries generated in the voice recognition dictionary, thereby updating the voice recognition dictionary.

20. A voice recognition dictionary generation apparatus according to claim 19, wherein, when the total number of the content dictionaries is equal to or larger than a predetermined limit, the control unit deletes the content dictionaries corresponding to the external device from the voice recognition dictionary, extracts all texts subjected to the voice recognition corresponding to the external device from the storage unit, generates new content dictionaries for the all texts corresponding to the external device stored in the storage unit, and stores the new content dictionaries generated in the voice recognition dictionary, thereby updating the voice recognition dictionary.