VOICE RECOGNITION DEVICE, VOICE RECOGNITION METHOD, AND VOICE RECOGNITION PROGRAM

• US 20100070277A1
• Filed: 02/26/2008
• Published: 03/18/2010
• Est. Priority Date: 02/28/2007
• Status: Active Grant

• Current Assignee: NEC Corporation
• Original Assignee: NEC Corporation
• Inventors: Tsujikawa, Masanori, Arakawa, Takayuki, Hanazawa, Ken
• Granted Patent: US 8,612,225 B2
• Time in Patent Office: Days
• US Class Current: 704/246
• CPC Class Codes: G10L 15/02 Feature extraction for speech recognition; Selection of recognition unit

Abstract

A voice recognition device that recognizes a voice of an input voice signal, comprises a voice model storage unit that stores in advance a predetermined voice model having a plurality of detail levels, the plurality of detail levels being information indicating a feature property of a voice for the voice model; a detail level selection unit that selects a detail level, closest to a feature property of an input voice signal, from the detail levels of the voice model stored in the voice model storage unit; and a parameter setting unit that sets parameters for recognizing the voice of an input voice according to the detail level selected by the detail level selection unit.

Claims

1. A voice recognition device that recognizes a voice of an input voice signal, comprising:

a voice model storage unit that stores in advance a predetermined voice model having a plurality of detail levels, said plurality of detail levels being information indicating a feature property of a voice for said voice model;

a detail level selection unit that selects a detail level, closest to a feature property of an input voice signal, from the detail levels of the voice model stored in said voice model storage unit; and

a parameter setting unit that sets parameters for recognizing the voice of an input voice according to the detail level selected by said detail level selection unit.

2. The voice recognition device as defined by claim 1, wherein said detail level selection unit finds a detail level on a unit time basis and selects the detail level closest to the feature property of the input voice signal.

3. The voice recognition device as defined by claim 2, wherein said detail level selection unit performs statistical analysis of the detail level, selected on a unit time basis, for a plurality of unit times and finds a detail level of one particular unit time of interest.

4. The voice recognition device as defined by claim 1, further comprising:

a distance calculation unit that calculates distance information indicating a difference between the voice model stored in said voice model storage unit and the feature property of an input voice signal wherein

said distance calculation unit calculates the distance information sequentially from low detail level distance information to higher detail level distance information, or sequentially from high detail level distance information to lower detail level distance information and

said detail level selection unit finds a detail level corresponding to a minimum of the distance information calculated by said distance calculation unit.

5. The voice recognition device as defined by claim 1, wherein said voice model storage unit stores in advance a voice model having a parent-child structure.

6. The voice recognition device as defined by claim 1, further comprising:

an acoustic model storage unit that stores a predetermined acoustic model in advance; and

a word string search unit that searches for, and extracts, a word string as a result of voice recognition based on the parameters that are set by said parameter setting unit, wherein

said acoustic model storage unit stores in advance an acoustic model having predetermined relevance to the voice model stored in said voice model storage unit; and

said word string search unit searches for, and extracts, a word string using relevance between the voice model and the acoustic model.

7. The voice recognition device as defined by claim 1, wherein said parameter setting unit sets at least one of a language weight parameter and a pruning parameter for performing predetermined pruning processing according to the detail level selected by said detail level selection unit.

8. The voice recognition device as defined by claim 1, further comprising:

an acoustic model storage unit that stores a plurality of predetermined acoustic models in advance;

a language model storage unit that stores a plurality of predetermined language models in advance; and

a model selection unit that selects a set of an acoustic model and a language model from the plurality of acoustic models, stored in said acoustic model storage unit, and the plurality of language models, stored in said language model storage unit, according to the detail level selected by said detail level selection unit.

9. The voice recognition device as defined by claim 1, further comprising:

an output change unit that changes an output method or an output content of a voice recognition result of the input voice signal according to the detail level selected by said detail level selection unit.

10. The voice recognition device as defined by claim 1, further comprising a voice model update unit that updates the voice model, stored in said voice model storage unit, according to the detail level selected by said detail level selection unit.

11. A voice recognition method that recognizes a voice of an input voice signal, comprising:

a detail level selection step that selects a detail level, closest to a feature property of an input voice signal, from a plurality of detail levels of a predetermined voice model stored in advance, said plurality of detail levels being information indicating a feature property of a voice for the voice model; and

a parameter setting step that sets parameters for recognizing the voice of an input voice according to the selected detail level.

12. The voice recognition method as defined by claim 11, wherein said detail level selection step finds a detail level on a unit time basis and selects the detail level closest to the feature property of the input voice signal.

13. The voice recognition method as defined by claim 12, wherein said detail level selection step performs a statistical analysis of the detail level, selected on a unit time basis, for a plurality of unit times and finds a detail level of one particular unit time of interest.

14. The voice recognition method as defined by claim 11, further comprising:

a distance calculation step that calculates distance information indicating a difference between the voice model stored in advance and the feature property of an input voice signal, wherein

said distance calculation step calculates the distance information sequentially from low detail level distance information to higher detail level distance information, or sequentially from high detail level distance information to lower detail level distance information and

said detail level selection step finds a detail level corresponding to a minimum of the calculated distance information.

15. The voice recognition method as defined by claim 11, wherein the detail level closest to the feature property of the input voice signal is selected based on a voice model stored in advance and having a parent-child structure.

16. The voice recognition method as defined by claim 11, further comprising:

a word string search step that searches for, and extracting, a word string as a result of voice recognition based on the parameters that are set, wherein

an acoustic model having predetermined relevance to the voice model is stored in advance, and

said word string search step searches for, and extracts, a word string using relevance between the voice model and the acoustic model.

17. The voice recognition method as defined by claim 11, wherein said parameter setting step sets at least one of a language weight

parameter and a pruning parameter for performing predetermined pruning processing according to the selected detail level.

18. The voice recognition method as defined by claim 11, further comprising a model selection step that selects a set of an acoustic model and a language model from a plurality of acoustic models stored in advance and a plurality of language models stored in advance according to the selected detail level.

19. The voice recognition method as defined by claim 11, further comprising an output change step that changes an output method or an output content of a voice recognition result of the input voice signal according to the selected detail level.

20. (canceled)

21. A voice recognition program that recognizes a voice of an input voice signal, said program causing a computer to execute:

detail level selection processing that selects a detail level, closest to a feature property of an input voice signal, from a plurality of detail levels of a predetermined voice model stored in advance, said plurality of detail levels being information indicating a feature property of a voice for said voice model; and

parameter setting processing that sets parameters for recognizing the voice of an input voice according to the selected detail level.

22-30. -30. (canceled)