Quality assessment apparatus and method

• US 20050015245A1
• Filed: 06/22/2004
• Published: 01/20/2005
• Est. Priority Date: 06/25/2003
• Status: Active Grant

• Current Assignee: Psytechnics Ltd. (NetScout Systems Incorporated)
• Original Assignee: Psytechnics Ltd. (NetScout Systems Incorporated)
• Inventors: Rix, Antony William, Goldstein, Tom, Barrett, Paul Alexander
• Granted Patent: US 7,412,375 B2
• Time in Patent Office: Days
• US Class Current: 704/226
• CPC Class Codes: G10L 25/69 for evaluating synthetic or decoded voice signals
• CPC Class Codes: H04M 3/2236 Quality of speech transmission monitoring

Abstract

This invention relates to a speech quality assessment system. The invention provides a method and apparatus for assessing the perceptual quality of stereo speech signals transmitted via a telecommunications network and recorded acoustically from an acoustic terminal device in which a mono reference signal comprising a single channel is aligned with a degraded stereo signal comprising a left and a right channel; a delay between each channel of said degraded signal and said reference signal is estimated; a noise masking indicator in dependence upon said estimated delays is generated; the level of the stereo signals is adjusted in dependence upon said noise masking indicator; a set of perceptually relevant parameters for each of said reference and degraded signals are generated; the perceptually relevant parameters of the reference signal with the perceptually relevant parameters of the degraded signal to generate a disturbance profile are compared; and a speech quality prediction is generated in dependence upon said disturbance profile.

Claims

1. An apparatus for assessing the perceptual quality of speech signals transmitted via a telecommunications network and recorded acoustically from an acoustic terminal device comprising:

a front end processor for aligning a mono reference signal comprising a single channel with a degraded stereo signal comprising a first channel and a second channel, said front end processor comprising a leveller for adjusting the power levels of said signals and a time aligner for determining the estimated delays for each of said channels of said degraded signal;

an auditory transformer for generating a set of perceptually relevant parameters for each of said signals; and

a comparator for comparing said perceptually relevant parameters to generate disturbance profiles; and

a modeller for generating a speech quality prediction in dependence upon said disturbance profiles;

in which said front end processor further comprises

a noise masking determiner for comparing signal parameters of each of said channels of said degraded signal and generating a noise masking indicator in dependence upon said parameters; and

in which said disturbance profiles are dependent upon said noise masking indicator.

2. An apparatus according to claim 1 in which said leveller adjusts the level of said signals in dependence upon whether noise masking is indicated.

3. An apparatus according to claim 1, in which the comparator is arranged to receive the noise masking indicator and in which the comparator is arranged to modify a disturbance profile in dependence upon a comparison between a disturbance profile for one channel and a set of perceptually relevant parameters for another channel when noise masking is indicated.

4. An apparatus according to claim 3, in which the comparator is arranged to received a voice activity signal and in which the disturbance profile is modified in dependence upon said voice activity signal.

5. An apparatus according to claim 1, in which said estimated delays comprise said signal parameters.

6. An apparatus according to claim 5, in which said noise masking determiner further comprises means for receiving an estimate of the confidence that each of said estimated delays is correct, and in which said noise masking indicator is generated in further dependence upon said estimated confidences.

7. An apparatus according to claim 1, in which said leveller is arranged

a) to adjust the level of each of said channels of the degraded signal in dependence upon only one channel of the signal when noise masking is indicated;

and

b) to adjust the level of each of said channels of the degraded signal independently when noise masking is not indicated.

8. An apparatus according to claim 7, in which said leveller is arranged to adjust the level of both channels in order to achieve a first predetermined RMS power level for said one channel at step a) and in which said levelling means is arranged to adjust the level of each of said channels to achieve a second predetermined RMS power level for both channels at step b).

9. An apparatus according to claim 8, in which said second predetermined level is greater than said first predetermined level.

10. A method of assessing the perceptual quality of stereo speech signals transmitted via a telecommunications network and recorded acoustically from an acoustic terminal device comprising the steps of:

aligning a mono reference signal comprising a single channel with a degraded stereo signal comprising a first channel and a second channel;

estimating a delay between each channel of said degraded signal and said reference signal;

generating a noise masking indicator in dependence upon a comparison of corresponding signal parameters for each channel;

generating a set of perceptually relevant parameters for each of said reference and degraded signals;

comparing said perceptually relevant parameters of the reference signal with the perceptually relevant parameters of the degraded signal to generate disturbance profiles; and

generating a speech quality prediction in dependence upon said disturbance profiles;

wherein said generated disturbance profiles are dependent upon said noise masking indicator.

11. A method according to claim 10, further comprising the stop of adjusting the level of the degraded signals in dependence upon said noise masking indicator.

12. A method according to claim 10, in which the comparing step comprises the sub-step of;

modifying a disturbance profile in dependence upon a comparison between a disturbance profile for one channel and a set of perceptually relevant parameters for the other channel when noise masking is indicated by said noise masking indicator.

13. A method according to claim 12, in which said modifying step is performed in dependence upon a voice activity signal.

14. A method according to claim 10, in which said estimated delays comprise said signal parameters.

15. A method according to claim 14, further comprising the step of estimating a confidence that each of said estimated delays is correct and generating the noise masking indicator (53) in dependence thereon.

16. A method according to claim 10, further comprising the steps of:

c) adjusting the level of each of said channels of the degraded signal in dependence upon only one channel of the signal when noise masking is indicated; and

d) adjusting the level of each of said channels of the degraded signal when noise masking is not indicated.

17. A method according to claim 16, in which step c) comprises adjusting the level of both channels in order to achieve a first predetermined RMS power level for one channel and in which step d) comprises adjusting the level of both channels independently to achieve a second predetermined RMS power level for both channels.

18. A method according to claim 17, in which the first predetermined level is greater than the second predetermined level.

19. A computer readable medium carrying a computer program for implementing the method according to claim 10.

20. A computer program for implementing the method according to claim 10.