Speech signal processing system, speech signal processing method and speech signal processing method program using noise environment and volume of an input speech signal at a time point

• US 8,793,128 B2
• Filed: 02/03/2012
• Issued: 07/29/2014
• Est. Priority Date: 02/04/2011
• Status: Active Grant

• Current Assignee: NEC Corporation
• Original Assignee: NEC Corporation
• Inventors: Miki, Kiyokazu
• Primary Examiner(s): ROBERTS, SHAUN A
• Application Number:
  US13/365,848
• Publication Number:
  US 20120271630A1
• Time in Patent Office: 907 Days
• Field of Search: 704/226, 704/233, 704/258, 704/260
• US Class Current: 704/233
• CPC Class Codes: G10L 13/033 Voice editing, e.g. manipulating the voice of the synthesiser
• CPC Class Codes: G10L 21/003 Changing voice quality, e.g. pitch or formants
• CPC Class Codes: G10L 25/84 for discriminating voice from noise

Abstract

A speech signal processing system that includes a speech input unit for inputting a speech signal; input speech storage unit for storing an input speech signal that is the speech signal inputted through the speech input unit; characteristic estimation unit for referring to the input speech signal stored in the input speech storage unit, and estimating characteristics of an input speech indicated by the input speech signal, the characteristics including an environmental sound included in the input speech signal; reference speech output unit for causing a predetermined speech signal that becomes a reference speech, to output; and characteristic adding unit for adding the characteristics of the input speech estimated by the characteristic estimation unit, in a reference speech signal that is the speech signal caused to output by the reference speech output unit.

Claims

1. A speech signal processing system comprising:

an input speech storage that stores an input speech signal;

a characteristic estimation unit that refers to the input speech signal stored in the input speech storage, and estimates characteristics of the input speech, the characteristics including an environmental sound included in the input speech signal, and the SN estimation unit obtains powers of the non-speech portion of the speech and a speech portion of the inputted speech signal and estimates an SN ratio;

a reference speech output that causes a predetermined speech signal that becomes a reference speech to be output;

a volume adjustment unit that adjusts the volume of the reference speech according to the SN ratio;

a characteristic adding unit that adds the estimated characteristics of the input speech, to the output reference speech signal of volume adjustment unit.

2. The speech signal processing system according to claim 1, wherein

the characteristic estimation unit estimates the environmental sound to be superimposed on a speech, the characteristics of the input speech based on at least one of a large amount of the speech signal, a small amount of the speech signal, and the absence of the speech signal.

3. The speech signal processing system according to claim 1, wherein

the characteristic adding unit emphasizes the estimated characteristics of the input speech, and adds the estimated characteristics of the input speech that have been emphasized, the reference speech signal.

4. The speech signal processing system according to claims 1, comprising:

response speech output unit outputs the signal output by the characteristic adding unit as a response speech signal.

7. An automatic speech response system comprising;

the speech signal processing system of claim 1;

a speech recognition unit which performs a speech recognition process for the input speech signal in the input speech storage;

a recognition result interpretation unit which extracts meaningful information from recognition result text outputted from the speech recognition unit and

a response speech generation unit which generated a response speech from a result of interpretation by the recognition result interpretation unit.

8. A speech recognition system having a diagnosis function comprising;

the speech signal processing system of claim 1;

a speech having known utterance content occurrence unit which causes a speech whose utterance content is known, to output as the reference speech;

a speech recognition unit which performs the speech recognition process for the speech signal in the input speech storage;

an acoustic environment determination unit which compares a result of the recognition of a converted reference speech by the speech recognition unit with the utterance content of the reference speech generated by the speech having known utterance content output unit, to obtain a recognition rate for the converted reference speech.

9. The speech recognition system according to claim 8, wherein

the acoustic environment determination unit determines whether the acoustic environment of the input speech is suitable for speech recognition based on a result of the comparison between a recognition result for a converted reference speech and the utterance content of the reference speech that is the speech having the known utterance content.

10. The speech recognition system according to claim 9, wherein

a result of a determination of whether the acoustic environment of the input speech is suitable is used in determination of whether the speech recognition result is acceptable, and for notifying the user to change the location or time and perform the input again.

5. A speech signal processing method comprising:

storing an input speech signal;

referring to the stored input speech signal;

estimating characteristics of an input speech indicated by the input speech signal, the characteristics including an environmental sound included in the input speech signal;

obtaining powers of the non-speech portion of the speech and a speech portion of the inputted speech signal and estimating an SN ratio;

causing a predetermined speech signal that becomes a reference speech to be output;

adjusting the volume of the reference speech according to the SN ratio; and

adding the estimated characteristics of the input speech, to the output reference speech signal.

6. A non-transitory computer readable storage medium storing a speech signal processing program to execute a method for causing a computer comprising an input speech storage unit to store an input speech signal that is an inputted speech signal, the method comprising:

storing an input speech signal;

referring to the stored input speech signal;

estimating characteristics of an input speech indicated by the input speech signal, characteristics including an environmental sound included in the input speech signal;

obtaining powers of the non-speech portion of the speech and a portion of the inputted speech signal and estimating an SN ratio;

causing a predetermined speech signal that becomes a reference speech to be output

adjusting the volume of the reference speech according to the SN ratio; and

adding the estimated characteristics of the input speech, to the output reference speech signal of volume adjustment unit.