Granted Patent B2

US 10,218,327 · App. 13/374,746 · Granted Feb 26, 2019

Abstract

Claims (65)

1. A communications apparatus comprising:

a receiving path over which received speech signals traverse in an audio stream, the audio stream including the received speech signals;

a processor including a dynamic audio enhancement device disposed in the receiving path to dynamically enhance the audio stream, the dynamic audio enhancement device configured to modify an amount of volume of the audio stream and an amount of equalization of the audio stream and comprising:

a noise level estimator (“NLE”) configured to generate a signal representing a noise level estimate, the noise level estimator including:

a non-stationary noise detector configured to detect non-stationary noise, and

a stationary noise detector configured to detect stationary noise,

wherein the noise level estimator is configured to generate as the signal a first signal representing a first noise level estimate, the generation of which is based on detection of the non-stationary noise,

wherein the noise level estimator is configured to generate as the signal a second noise level estimate, the generation of which is based on detection of the stationary noise,

wherein the dynamic audio enhancement device is configured to modify the amount of the volume to generate a modified amount of volume based on the signal representing the noise level estimate to form a modified audio stream, and is further configured to modify the amount of the equalization to generate a modified amount of equalization based on the signal representing the noise level estimate to form the modified audio stream,

wherein the modified audio stream is used to generate audio at a loudspeaker.

2. The apparatus of claim 1 wherein the noise level estimator further comprises:

a bi-modal noise level estimator configured to generate noise level estimates differently based on the detection of different types of noise.

3. The apparatus of claim 1 further comprising:

an automatic volume control (“AVC”) configured to implement either the first noise level estimate or the second noise level estimate.

4. The apparatus of claim 1 wherein the received speech signals comprise:

a signal (“O 2 ”) generated by an omnidirectional microphone.

5. The apparatus of claim 1 wherein the non-stationary noise detector comprises:

a nonlinear peak tracking filter configured to receive a signal representing speech-like non-stationary noise, and further configured to generate a filtered signal including a largest absolute value of the speech-like non-stationary noise in a frame; and

a low pass filter configured to smooth the filtered signal.

6. The apparatus of claim 1 wherein the stationary noise detector comprises:

an energy computation module configured to compute the energy of a one of the received speech signals; and

a moving average filter.

7. The apparatus of claim 1 wherein the dynamic audio enhancement device further comprises:

a limiter configured to introduce a frame of delay to detect a gain to be applied to the received speech signals, and to attenuate a magnitude of the received speech signals.

8. The apparatus of claim 7 wherein the limiter is further configured to prevent clipping at a speaker by attenuating the magnitude of the received speech signals.

9. The apparatus of claim 7 wherein the limiter is further configured to reduce strongly-voiced portions of the received speech signals to amplify unvoiced and weakly-voiced portions of the received speech signals.

10. The apparatus of claim 1 further comprises:

either a headset or a handset.

11. A method comprising:

receiving an audio stream via a receiving processing path, the audio stream including received speech signals;

detecting a type of noise at a processor;

generating a signal representing either a first noise level estimate, the generation of which is based on detection of non-stationary noise as the type of noise, or a second noise level estimate, the generation of which is based on detection of stationary noise as the type of noise;

modifying an amount of volume of the audio stream to form a modified amount of volume,

wherein the modified amount of volume is based on either the first noise level estimate or the second noise level estimate to reduce or eliminate environmental acoustic noise associated with the received speech signals;

modifying an amount of equalization of the audio stream to form a modified amount of equalization,

wherein the modified amount of equalization is based on either the first noise level estimate or the second noise level estimate; and

implementing either the modified amount of volume or the modified amount of equalization, or both, to form a modified audio stream to generate audio at a loudspeaker.

12. The method of claim 11 further comprising:

introducing a delay to form a delayed portion of the received speech signals;

detecting a gain to be applied to the delayed portion of the received speech signals at a first point in time; and

attenuating a magnitude of one of the received speech signals at a second point in time.

13. The method of claim 12 wherein attenuating the magnitude of the one of the received speech signals further comprises:

preventing clipping at a speaker.

14. The method of claim 12 wherein attenuating the magnitude of the one of the received speech signals further comprises:

reducing magnitudes of strongly-voiced portions of the received speech signals; and

amplifying unvoiced and weakly-voiced portions of the received speech signals.

15. The method of claim 11 further comprising:

applying the first noise level estimate or the second noise level estimate to an automatic volume control (“AVC”); and

applying the first noise level estimate or the second noise level estimate to an equalizer.

16. The method of claim 11 wherein receiving received speech signals comprises:

receiving a signal (“O 2 ”) generated by an omnidirectional microphone.

17. The method of claim 11 further comprising:

identifying a first mode of operation;

identifying in a frame a largest absolute value of a received speech signal, the received speech signal being one of the received speech signals;

filtering nonlinearly the largest absolute value of the received speech signal to form a resultant signal; and

applying the resultant signal to a low pass filter.

18. The method of claim 11 further comprising:

identifying a second mode of operation;

computing the energy of a one of the received speech signals; and

applying the one of the received speech signals to a moving average filter.

19. The method of claim 11 further comprising:

receiving a compressor gain; and

compressing high speech levels and boosting low speech levels.

20. The method of claim 11 further comprising:

adding wind comfort noise.

Assignments (17)