Small array microphone apparatus and noise suppression methods thereof

1. Field of the Invention

The invention relates to a small array microphone, and in particular to noise suppression using small array microphone.

2. Description of the Related Art

Noise suppression is often required in many communication systems and voice recognition devices to suppress noise to improve communication quality and voice recognition performance. Noise suppression may be achieved using various techniques, which may be classified as single microphone techniques and array microphone techniques.

Array microphone noise reduction technique uses multiple microphones placed at different locations and separated from each other by some minimum distance to form a beam. Conventionally, the beam is used to pick up speech that is then used to reduce the amount of noise picked up outside the beam. Thus, the array microphone techniques can suppress non-stationary noise. Multiple microphones, however, also themselves create more noise.

Thus, effective suppression of noise in communication system and voice recognition devices is desirable.

A detailed description is given in the following embodiments with reference to the accompanying drawings.

An embodiment of a small array microphone apparatus is provided. The small array microphone apparatus comprises first and second omni-directional microphones, a microphone calibration unit and a directional microphone forming unit. The first and second omni-directional microphones respectively convert sound from a desired near-end talker into first and second signals. The second and first omni-directional microphones and the desired near-end talker are arranged in a line. The microphone calibration unit receives the first and second signals, calibrates on gain, and correspondingly outputs first and second calibration signals. The directional microphone forming unit receives the first and second calibration signals to output a first directional microphone signal with a predefined directivity according to a control signal and a second directional microphone signal with a fixed directivity for noise detection. Establishment of the control signal is based on whether environmental noise power generated by an environmental detection unit exceeds a predefined threshold.

An embodiment of a noise suppression method is provided. The noise suppression method comprises arranging first and second omni-directional microphones and a desired near-end talker in a line, calibrating each band of a first signal and second signal from the first and second omni-directional microphones to correspondingly generate first and second calibration signals, generating a first directional microphone signal with a predefined directivity according to the first calibration signal, the second calibration signal, and a control signal, and generating a second directional microphone signal with fixed directivity for noise detection according to the first and second calibration signals. Determination of the control signal is based on whether environmental noise power exceeds a predefine threshold.

The present invention can be more fully understood by reading the subsequent detailed description and examples with references made to the accompanying drawings, wherein:

FIG. 1 is a schematic diagram of a small array microphone apparatus according to an embodiment of the invention;

FIG. 2 is a schematic diagram of a microphone calibration unit according to another embodiment of the invention;

FIG. 3 is a schematic diagram of a directional microphone forming unit according to another embodiment of the invention;

FIG. 4 is a schematic diagram of a detection unit according to another embodiment of the invention;

FIG. 5 is a schematic diagram of a small array microphone apparatus according to another embodiment of the invention; and

FIG. 6 is a schematic diagram of a directional microphone forming unit according to another embodiment of the invention.

The following description is of the best-contemplated mode of carrying out the invention. This description is made for the purpose of illustrating the general principles of the invention and should not be taken in a limiting sense. The scope of the invention is best determined by reference to the appended claims.

FIG. 1 is a schematic diagram of small array microphone apparatus 100 according to an embodiment of the invention. Small array microphone apparatus 100 comprises omni-directional microphones Mic 1 and Mic 2 , microphone calibration unit 110 , directional microphone forming unit 120 , time domain noise suppression unit 130 , adaptive channel forming unit 140 , transformer 150 , detection unit 155 , frequency domain noise suppression unit 180 , SNR based equalizer 185 and inverse transformer 190 . Small array microphone apparatus 100 detects environmental noise to adjust directional microphone signals dm 1 and dm 2 of directivity for noise suppression. In addition, detection unit 155 comprises ambient noise estimate unit 160 and environmental detection unit 170 .

As shown in FIG. 1 , the desired near-end talker P 1 and omni-directional microphone Mic 1 and Mic 2 are arranged in a line, referred to as an end-fire way. Omni-directional microphone Mic 1 and Mic 2 respectively convert sound from the desired near-end talker 10 into signals S 1 and S 2 . Microphone calibration unit 110 receives signals S 1 and S 2 , calibrates on gain, and correspondingly outputs calibration signals C 1 and C 2 . Directional microphone forming unit 120 receives calibration signals C 1 and C 2 and outputs directional microphone signal dm 1 with a predefined directivity according to control signal Ctrl and directional microphone signal dm 2 with a fixed directivity for noise detection. Control signal Ctrl is determined by whether environmental noise power generated by environmental detection unit 170 exceeds a predefined threshold. According to another embodiment of the invention, the directional microphone signal dm 2 with the fixed directivity is a signal with a cardioid, super-cardioid or hyper-cardioid polar pattern for noise detection. The directional microphone signal dm 1 with predefined directivity is a signal with a similar omni-directional polar pattern when the environmental noise power is below the predefined threshold. The directional microphone signal dm 1 with predefined directivity is a signal with a cardioid, super-cardioid or hyper-cardioid polar pattern when the environmental noise power exceeds the predefined threshold.

Time domain noise suppression unit 130 receives directional microphone signals dm 1 and dm 2 and calibration signal C 2 , suppresses noise, and correspondingly outputs directional signals d 1 and d 2 and calibration signal C 3 to adaptive channel forming unit 140 .

Adaptive channel forming unit 140 receives directional signals d 1 and d 2 and calibration signal C 3 to respectively generate first main channel signal m 1 , second main channel signal m 2 and reference channel signal r 1 . Second main channel signal m 2 is indirectly provided to ambient noise estimate unit 160 for environmental detection.

Transformer 150 transforms first main channel signal m 1 , second main channel signal m 2 and reference signal r 1 from time domain to frequency domain to correspondingly output main channel signals M 1 and M 2 and reference channel signal R 1 . Main channel signal M 2 and reference channel R 1 , frequency domain signals, are provided to ambient noise estimate unit 160 of detection unit 155 .

Ambient noise estimate unit 160 receives and compares reference channel signal R 1 and main channel signal M 2 to output control signals Co 1 and Co 2 and noise estimate signal N 1 to environmental detection unit 170 . Environmental detection unit 170 generates control signal Ctrl according to control signals Co 1 and Co 2 and noise estimate signal N 1 to control directional microphone signal dm 1 with the predefined directivity.

Frequency domain noise suppression unit 180 receives main channel signal M 1 and noise estimate signal N 1 , suppresses noise of main channel signal M 1 according to noise estimate signal N 1 and generates clear voice signal V 1 . SNR based equalizer 185 equalizes clear voice signal V 1 to generate clear voice signal V 2 . Inverse transformer 190 transforms clear voice signal V 2 from frequency domain to time domain to generate clear voice signal v 2 .

FIG. 2 is a schematic diagram of microphone calibration unit 110 according to another embodiment of the invention. Microphone calibration unit 110 comprises power detection unit 112 , power smoothing unit 114 , calibration unit 116 and subband synthesis unit 118 . Power detection unit 112 comprises subband analysis unit 1121 , power calculation in all bands unit 1122 and voice activity detection unit 1123 . Power detection unit 112 detects power of each band of signals S 1 and S 2 . Power smoothing unit 114 smoothes each band of signals S 1 and S 2 . Calibration unit 116 comprises calibrating gains for all bands unit 1161 and applying mic gains for all bands unit 1162 . Calibrating gains for all bands unit 1161 calibrates each band of signals S 1 and S 2 by multiplying calibrating gains to each band of the signal S 1 , wherein the calibrating gains are generated by each band of signal S 2 divided by each band of signal S 1 . Applying gains for all bands unit 1162 may comprise multiplication of a predefined gain for all bands of signals S 1 and S 2 . Subband synthesis unit 118 synthesizes each band of signals S 1 and S 2 to generate calibration signals X 1 and X 2 .

FIG. 3 is a schematic diagram of directional microphone forming unit 120 according to another embodiment of the invention. Directional microphone forming unit 120 comprises first phase adjustment unit 121 , second phase adjustment unit 122 , fixed phase adjustment unit 123 , and subtractors 124 and 125 .

First phase adjustment unit 121 shifts calibration signal X 1 first phase P 1 according to control signal Ctrl to generate signal XP 1 . First phase P 1 is a positive value P 0 for compensating sound propagation from omni-directional microphone Mic 1 to omni-directional microphone Mic 2 when the environmental noise power is below the predefined threshold. Phase P 1 is less than the positive value P 0 when the environmental noise power exceeds the predefined threshold. The environmental noise power is detected by detection device 155 .

Second phase adjustment unit 122 shifts calibration signal X 2 second phase P 2 according to control signal Ctrl to generate signal XP 2 . Second phase P 2 is 180° for two calibration signal X 1 and X 2 added together with the same phase when the environmental noise power is below the predefined threshold. Second phase P 2 is 0° when the environmental noise power exceeds the predefined threshold.

Fixed phase adjustment unit 123 shifts calibration signal X 2 fixed phase P 3 to generate signal XP 3 . First subtractor 124 subtracts signal XP 2 from signal XP 1 to generate first directional microphone signal dm 1 , directivity of which is changed by control signal Ctr 1 . Second subtractor 125 subtracts signal XP 3 from signal X 1 to generate the second directional microphone signal dm 2 with fixed directivity, such as super-cardioid or hyper-cardioid for noise detection.

FIG. 4 is a schematic diagram of detection unit 155 according to another embodiment of the invention. Detection unit 155 comprises ambient noise estimate unit 160 and environmental detection unit 170 . Ambient noise estimate unit 160 comprises entire power calculating units 1621 and 1622 , each frequency bin power calculating units 1641 and 1642 , power smoothing units 1651 , 1652 , 1653 and 1654 , comparing units 1671 and 1672 and noise estimate unit 168 . Entire power calculating unit 1621 calculates the entire power of reference channel signal R 1 to output power signal Pw 1 . Power smoothing unit 1651 smoothes power signal Pw 1 to output power signal Ps 1 . Each frequency bin power calculating unit 1641 calculates the power of each frequency bin to output power signal Bw 1 . Power smoothing unit 1652 smoothes power signal Bw 1 to output power signal Bs 1 .

Similarly, entire power calculating unit 1622 calculates the entire power of main channel signal M 2 to output power signal Pw 2 . Power smoothing unit 1654 smoothes power signal Pw 2 to output power signal Ps 2 . Each frequency bin power calculating unit 1642 calculates the power of each frequency bin to output power signal Bw 2 . Power smoothing unit 1653 smoothes power signal Bw 2 to output power signal Bs 2 . It is noted that main channel signal M 2 provides noise detection.

Comparing unit 1672 compares power signals Ps 1 and Ps 2 to generate control signal Co 1 . Control signal Co 1 is power signal Ps 1 divided by power signal Ps 2 . Similarly, comparing unit 1671 compares power signals Bs 1 and Bs 2 to generate control signal Co 2 . Control signal Co 2 is power signal Bs 1 divided by power signal Bs 2 . Noise estimate unit 168 receives control signals Co 1 and Co 2 and power signal Bs 1 to generate noise estimate signal N 1 . Environmental detection unit 170 generates control signal Ctrl to control directional microphone unit 120 to form different polar patterns according to control signals Co 1 and Co 2 and power signal Bs 1 more or less than predefined values. If all control signals Co 1 and Co 2 and power signal Bs 1 are more than predefined values, it is determined that the environmental noise power exceeds the predefined threshold (noise environment) and the polar pattern of first directional microphone signal dm 1 is super-cardioid or hyper-cardioid polar pattern.

If none of control signals Co 1 and Co 2 and power signal Bs 1 exceeds predefined values, it means that the environmental noise power doesn't exceed the predefined threshold (quiet environment) and the polar pattern of first directional microphone signal dm 1 is a similar omni-directional polar pattern.

FIG. 5 is a schematic diagram of a small array microphone apparatus 500 according to another embodiment of the invention. Small array microphone apparatus 500 comprises omni-directional microphones Mic 1 , Mic 2 and Mic 3 , microphone calibration unit 510 , directional microphone forming unit 520 , time domain noise suppression unit 130 , adaptive channel forming unit 140 , transformer 150 , detection unit 155 , frequency domain noise suppression unit 180 , SNR based equalizer 185 and inverse transformer 190 . The differences between small array microphone apparatus 500 and small array microphone apparatus 100 are one more omni-directional microphones Mic 3 , microphone calibration unit 510 and directional microphone forming unit 520 . Especially, directional microphone forming unit 520 is big different and discussed as followed.

FIG. 6 is a schematic diagram of directional microphone forming unit 520 according to another embodiment of the invention. Directional microphone forming unit 520 comprises first phase adjustment unit 521 , second phase adjustment unit 522 , third phase adjustment unit 523 , fixed phase adjustment unit 524 , fifth phase adjustment unit 528 , sixth phase adjustment unit 529 and subtractors 525 , 526 and 527 . Directional microphone forming unit 520 is a two order directional microphone forming unit with two-stage processing. In the first stage, calibration signals X 1 , X 2 and X 3 are respectively sent to first phase adjustment unit 521 , second phase adjustment unit 522 and third phase adjustment unit 523 to phase-shift P 1 for calibration signal X 1 , P 2 for calibration signal X 2 and P 3 for calibration signal X 3 to acquire three phase shifted signals XP 1 , XP 2 and XP 3 . Subtractors 525 and 526 generate signals X 11 and X 21 by subtracting signal XP 2 from signal XP 1 and signal XP 3 from signal XP 2 . Control signal Ctrl is used to control the phase shift values, P 1 , P 2 and P 3 , to get three phase shifted signal XP 1 , XP 2 and XP 3 and further forms the first stage directivity. In the second stage, signals X 11 and X 21 are respectively sent to fifth phase adjustment unit 528 and sixth phase adjustment unit 529 to phase-shift P 11 for signal X 11 and P 21 for signal X 21 to get two phase shifted signals XP 4 and XP 5 .

Subtractor 531 generates first directional microphone signal dm 1 with a predefined directivity by subtracting signal XP 5 from signal XP 4 . Control signal Ctrl is used to control the phase shift values, P 11 and P 21 , to acquire two phase shifted signals XP 4 and XP 5 and further forms the second stage directivity. Similarly, subtractor 527 generates second directional microphone signal dm 2 with a fixed directivity by subtracting signal XP 4 from calibration signal X 2 .

While the invention has been described by way of example and in terms of preferred embodiment, it is to be understood that the invention is not limited thereto. To the contrary, it is intended to cover various modifications and similar arrangements (as would be apparent to those skilled in the art). Therefore, the scope of the appended claims should be accorded the broadest interpretation so as to encompass all such modifications and similar arrangements.