SPEECH RECOGNITION DEVICE

• US 20150100317A1
• Filed: 01/29/2013
• Published: 04/09/2015
• Est. Priority Date: 04/16/2012
• Status: Active Grant

• Current Assignee: DENSO Corporation
• Original Assignee: DENSO Corporation
• Inventors: Tsuji, Hideaki, Miyaguni, Satoshi
• Granted Patent: US 9,704,479 B2
• Time in Patent Office: Days
• US Class Current: 704/243
• CPC Class Codes: G10L 15/06 Creation of reference templates; Training of speech recognition systems, e.g. adaptation to the characteristics of the speaker's voice G10L15/14 takes precedence

Abstract

A speech recognition device starts to generate dictionary data for each type of name based on name data and paraphrase data, and executes dictionary registration of the dictionary data. The speech recognition device obtains text information same as text information for generating the dictionary data last time. When back-up data corresponding to the last time text information is generated, the speech recognition device executes the dictionary registration of the dictionary data generated as the back-up data. Further, a dictionary data generation device executes the dictionary registration of the dictionary data based on given name data every time the dictionary data generation device completes generation of the dictionary data based on the given name data.

Claims

1. A speech recognition device comprising:

a text information acquisition device that acquires text information including a plurality of name data from an external device;

a dictionary data generation device that generates dictionary data based on the plurality of name data included in the text information;

a dictionary data saving control device that controls a dictionary data saving device (15, 16) to store the dictionary data;

a dictionary registration device that registers the dictionary data into a dictionary;

a speech recognition device that performs speech recognition using the dictionary data registered in the dictionary; and

a backup data generation device that generates backup data of the dictionary data stored in the dictionary data saving device,

wherein;

the dictionary data generation device generates the dictionary data for each type of name based on the name data; and

in a case where text information same as text information acquired from the external device at a present moment has been obtained last time from the external device, and backup data corresponding to the text information is already generated by the backup data generation device, the dictionary registration device registers the dictionary data generated as the backup data into the dictionary, and the dictionary registration device registers the dictionary data based on given name data into the dictionary every time the dictionary data generation device completes generation of dictionary data based on the given name data.

2. The speech recognition device according to claim 1, wherein:

the dictionary data generation device generates the dictionary data based on the name data in an order from higher to lower frequencies of use as the dictionary data when the speech recognition device has performed speech recognition in the past.

3. The speech recognition device according to claim 1, wherein:

the dictionary data generation device generates the dictionary data based on the name data in an order from smaller to larger numbers of subjects, for which the dictionary data is generated.

4. The speech recognition device according to claim 1, wherein:

when given name data is changed since the text information has been acquired last time, the dictionary data generation device generates dictionary data based on changed name data; and

when given name data is not changed since the text information has been acquired last time, the dictionary data generation device skips generation of dictionary data based on unchanged name data.

5. The speech recognition device according to claim 1, further comprising:

a correlation device that correlates a given name with paraphrase data,

wherein;

the dictionary data generation device generates dictionary data based on the paraphrase data correlated with the given name by the correlation device after generation of the dictionary data based on all of the name data is completed; and

when the dictionary data generation device completes generation of dictionary data based on the paraphrase data, the dictionary registration device registers the dictionary data based on the paraphrase data into the dictionary.

6. The speech recognition device according to claim 5, wherein:

when the correlation device respectively correlates a plurality of names with a plurality of paraphrase data, the dictionary data generation device generates dictionary data based on the plurality of paraphrase data for each type of name; and

every time the dictionary data generation device completes generation of dictionary data based on given paraphrase data, the dictionary registration device registers the dictionary data based on the given paraphrase data into the dictionary.

7. The speech recognition device according to claim 5, wherein:

the dictionary data generation device generates the dictionary data based on the plurality of paraphrase data in an order from higher to lower frequencies of use of the paraphrase data as the dictionary data when the speech recognition device has performed speech recognition in the past.

8. The speech recognition device according to claim 5, wherein:

the dictionary data generation device generates the dictionary data based on the plurality of paraphrase data in an order from smaller to larger numbers of subjects, for which the dictionary data is generated.

9. The speech recognition device according to claim 5, wherein:

when given paraphrase data is changed since the text information has been acquired last time, the dictionary data generation device generates dictionary data based on changed paraphrase data; and

when given paraphrase data is not changed since the text information has been acquired last time, the dictionary data generation device skips generation of dictionary data based on unchanged paraphrase data.

10. The speech recognition device according to claim 1, further comprising:

a notification control device that provides a notification using a notification device that speech recognition using the backup data by the speech recognition device is available when the backup data is already generated by the backup data generation device in a period, during which the dictionary data generation device is generating the dictionary data and the dictionary registration device is registering the dictionary data into the dictionary.

11. The speech recognition device according to claim 10, wherein:

when the dictionary data generation device completes generation of dictionary data based on the name data and the paraphrase data, the notification control device provides a notification using the notification device that speech recognition using the dictionary data by the speech recognition device is available.

12. The speech recognition device according to claim 1, further comprising:

a notification control device that provides a notification using the notification device after generation of dictionary data based on the name data and the paraphrase data is completed by the dictionary data generation device that speech recognition using the dictionary data by the speech recognition device is available when the backup data is not generated by the backup data generation device in a period, during which the dictionary data generation device is generating the dictionary data.

13. The speech recognition device according to claim 11, wherein:

the notification control device provides a notification using the notification device immediately after the dictionary data generation device completes generation of dictionary data based on the name data and the paraphrase data that speech recognition using the dictionary data by the speech recognition device is available.

14. The speech recognition device according to claim 1, wherein:

the text information acquisition device acquires music information including at least an artist name, an album name, a title name, and a playlist name as the text information.