Audio beamforming with nulling control system and methods

Abstract

Audio beamforming systems and methods that enable more precise control of lobes and nulls of an array microphone are provided. Optimized beamformer coefficients can be generated to result in beamformed signals associated with one or more lobes steered towards one or more desired sound locations and one or more nulls steered towards one or more undesired sound location. The performance of acoustic echo cancellation can be improved and enhanced.

Description

TECHNICAL FIELD

This application generally relates to audio beamforming. In particular, this application relates to audio beamforming systems and methods that enable more precise control of lobes and nulls of an array microphone, which can result in performance improvements related to acoustic echo cancellation of audio captured by the array microphone.

BACKGROUND

Conferencing environments, such as conference rooms, boardrooms, video conferencing applications, and the like, can involve the use of microphones for capturing sound from various audio sources that are active in such environments. Such audio sources may include humans talking, for example. The captured sound may be disseminated to a local audience in the environment through amplified speakers (for sound reinforcement), and/or to others remote from the environment (such as via a telecast and/or a webcast). The types of microphones and their placement in a particular environment may depend on the locations of the audio sources, physical space requirements, aesthetics, room layout, and/or other considerations. For example, in some environments, the microphones may be placed on a table or lectern near the audio sources. In other environments, the microphones may be mounted overhead to capture the sound from the entire room, for example. Accordingly, microphones are available in a variety of sizes, form factors, mounting options, and wiring options to suit the needs of particular environments.

Array microphones having multiple microphone elements can provide benefits such as steerable coverage or pick up patterns having lobes and/or nulls, which allow the microphones to focus on desired sound sources and reject unwanted sounds such as room noise and other undesired sound sources. The ability to steer audio pick up patterns provides the benefit of being able to be less precise in microphone placement, and in this way, array microphones are more forgiving. Moreover, array microphones provide the ability to pick up multiple audio sources with one array microphone or unit, again due to the ability to steer the pickup patterns.

Beamforming is used to combine signals from the microphone elements of array microphones in order to achieve a certain pickup pattern having one or more lobes and/or nulls. However, even though the lobes of a pickup pattern may be steered to detect sounds from desired sound sources (e.g., a talker in the local environment), the lobes may also detect sounds from undesired sound sources. The detection of sounds from undesired sound sources may be particularly exacerbated when a loudspeaker is in close physical proximity to the microphone elements of an array microphone. For example, the microphone elements may pick up the sound from a remote location (e.g., the far end of a teleconference) that is being played on the loudspeaker. In this situation, the audio transmitted to the remote location may therefore include an undesirable echo, e.g., sound from the local environment as well as sound from the remote location.

Acoustic echo cancellation systems may be able to remove such echo that is picked up by the array microphone before the audio is transmitted to the remote location. However, an acoustic echo cancellation system may work poorly and have suboptimal performance if it needs to constantly readapt and/or is overwhelmed, such as when the sound from a physically proximate loudspeaker is being continually detected by the array microphone. For example, the sound from the loudspeaker (which may include audio from the remote location) may not be completely cancelled by an acoustic echo cancellation system and may be transmitted to the remote location. This may result in overall decreased user satisfaction with the array microphone.

Accordingly, there is an opportunity for audio beamforming systems and methods that enable more precise control of lobes and nulls of an array microphone to optimize the performance of acoustic echo cancellation of audio captured by array microphones.

SUMMARY

The techniques of this disclosure are intended to solve the above-described problems by providing audio beamformer systems and methods that are designed to, among other things: (1) generate a set of beamformer coefficients that result in a beamformed signal associated with one or more lobes steered towards one or more desired sound locations and one or more nulls steered towards one or more undesired sound locations; (2) optimize the beamformer coefficient generation through the use of a process that may involve matrix inversion(s) and/or Lagrange multipliers; and (3) improve and enhance the performance of a downstream acoustic echo cancellation process by more precisely steering nulls to minimize the detection of undesired sound and/or by attenuating specific frequency ranges.

In an embodiment, a method includes receiving a plurality of audio signals from a plurality of microphones; receiving a first steering vector associated with a desired sound location and a second steering vector associated with an undesired sound location; generating a set of beamformer coefficients based on the first steering vector and the second steering vector, where the set of beamformer coefficients is generated to result in a beamformed signal associated with a lobe steered towards the desired sound location and a null steered towards the undesired sound location; and generating a beamformed signal based on the plurality of audio signals and the set of beamformer coefficients, using a frequency domain beamforming technique.

In another embodiment, an audio device includes a plurality of microphones configured to generate a plurality of audio signals, a coefficient generator, and a beamformer in communication with the plurality of microphones and the coefficient generator. The coefficient generator is configured to receive a first steering vector associated with a desired sound location and a second steering vector associated with an undesired sound location, and generate a set of beamformer coefficients based on the first steering vector and the second steering vector, where the set of beamformer coefficients is generated to result in a beamformed signal associated with a lobe steered towards the desired sound location and a null steered towards the undesired sound location. The beamformer is configured to generate a beamformed signal based on the plurality of audio signals and the set of beamformer coefficients, using a frequency domain beamforming technique.

These and other embodiments, and various permutations and aspects, will become apparent and be more fully understood from the following detailed description and accompanying drawings, which set forth illustrative embodiments that are indicative of the various ways in which the principles of the invention may be employed.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a block diagram of an audio beamforming with nulling control system for use with an array microphone, in accordance with some embodiments.

FIG. 2 is a flowchart illustrating operations for the beamforming of audio signals of a plurality of microphones using the audio beamforming with nulling control system of FIG. 1, in accordance with some embodiments.

FIG. 3 is a flowchart illustrating operations for the generation of beamformer coefficients, in accordance with some embodiments.

FIGS. 4A-4D are exemplary diagrams depicting a pickup pattern with a single lobe steered towards a desired sound source and a single null steered towards an undesired sound source; a pickup pattern with a single lobe steered towards the desired sound source and multiple nulls steered towards one or more undesired sound sources; and a pickup pattern with multiple lobes steered towards one or more desired sound sources and one or more nulls steered towards one or more undesired sound sources.

DETAILED DESCRIPTION

The description that follows describes, illustrates and exemplifies one or more particular embodiments of the invention in accordance with its principles. This description is not provided to limit the invention to the embodiments described herein, but rather to explain and teach the principles of the invention in such a way to enable one of ordinary skill in the art to understand these principles and, with that understanding, be able to apply them to practice not only the embodiments described herein, but also other embodiments that may come to mind in accordance with these principles. The scope of the invention is intended to cover all such embodiments that may fall within the scope of the appended claims, either literally or under the doctrine of equivalents.

It should be noted that in the description and drawings, like or substantially similar elements may be labeled with the same reference numerals. However, sometimes these elements may be labeled with differing numbers, such as, for example, in cases where such labeling facilitates a more clear description. Additionally, the drawings set forth herein are not necessarily drawn to scale, and in some instances proportions may have been exaggerated to more clearly depict certain features. Such labeling and drawing practices do not necessarily implicate an underlying substantive purpose. As stated above, the specification is intended to be taken as a whole and interpreted in accordance with the principles of the invention as taught herein and understood to one of ordinary skill in the art.

The audio beamforming systems and methods described herein can enable more precise control of lobes and nulls of an array microphone, which can result in performance improvements related to acoustic echo cancellation (AEC) processing of audio captured by the array microphone. The systems and methods may generate beamformer coefficients that can result in a beamformed signal associated with one or more lobes that are steered towards one or more desired sound locations, and also one or more nulls that are steered towards one or more undesired sound locations. The beamformer coefficients may be generated based on steering vectors associated with the locations of desired sounds (e.g., talkers) and the locations of undesired sounds (e.g., sound from a loudspeaker). In embodiments, a beamformer coefficient generation process may be implemented to optimize the coefficients through the use of a process involving matrix inversion and Lagrange multipliers.

Based on determining optimized beamformer coefficients, audio signals from the elements of the array microphone may be processed using a frequency domain beamforming technique and/or a time domain beamforming technique to generate a beamformed signal. The beamformed signal may include sound detected in the environment that is the result of more precise steering of nulls caused by the optimized beamformer coefficients and the ensuing minimization of the detection of undesired sound in an environment. The performance of a downstream AEC may accordingly be enhanced and improved since the beamformed signal may include less undesired sound, e.g., sound from a remote location that is played on a loudspeaker.

In embodiments, specific frequency ranges may be attenuated by a lobe and/or a null (such as voice rich frequency ranges). Such attenuation may also enhance the performance of a downstream AEC by removing sounds that the AEC no longer has to process. Through the use of the audio beamforming systems and methods described herein, latency and computational resources related to the AEC can be reduced, resulting in improved performance of the AEC. In addition, the occurrence of the undesirable echo of persons at a remote location hearing their own speech and sound can be reduced.

FIG. 1 is a block diagram of an audio beamforming with nulling control system 100 for use with an array microphone. The system 100 may include any suitable number of microphone elements 102a, b, c, . . . , z that are included in the array microphone, a coefficient generator 104, a beamformer 106, a mixer 108, and an acoustic echo canceller 110. Various components included in the system 100 may be implemented using software executable by a computing device with a processor and memory, and/or by hardware (e.g., discrete logic circuits, application specific integrated circuits (ASIC), programmable gate arrays (PGA), field programmable gate arrays (FPGA), etc. An exemplary embodiment of a method 200 for beamforming with nulling control, mixing, and echo cancelling of audio signals using the system 100 is described in more detail below with reference to FIG. 2.

The array microphone that includes the microphone elements 102a, b, c, . . . , z can detect sounds from audio sources at various frequencies. The array microphone may be utilized in a conference room or boardroom, for example, where the audio sources may be one or more human talkers and/or other desirable sounds. Other sounds may be present in the environment which may be undesirable, such as sounds from loudspeakers (e.g., sound from a remote location of a teleconference), noise from ventilation, other persons, audio/visual equipment, electronic devices, etc. In a typical situation, the audio sources may be seated in chairs at a table, although other configurations and placements of the audio sources are contemplated and possible.

The array microphone may be placed on a table, lectern, desktop, etc. so that the sound from the audio sources can be detected and captured, such as speech spoken by human talkers. The array microphone may be able to form multiple pickup patterns using the system 100 so that the sound from the audio sources is more consistently detected and captured. The microphone elements 102a, b, c, . . . , z may be arranged in any suitable layout, including in concentric rings and/or be harmonically nested. The microphone elements 102a, b, c, . . . , z may be arranged to be generally symmetric or may be asymmetric, in embodiments. In further embodiments, the microphone elements 102a, b, c, . . . , z may be arranged on a substrate, placed in a frame, or individually suspended, for example.

The microphone elements 102a, b, c, . . . , z may each be a MEMS (micro-electrical mechanical system) microphone, in some embodiments. In other embodiments, the microphone elements 102a, b, c, . . . , z may be electret condenser microphones, dynamic microphones, ribbon microphones, piezoelectric microphones, and/or other types of microphones. In embodiments, the microphone elements 102a, b, c, . . . , z may be unidirectional microphones that are primarily sensitive in one direction. In other embodiments, the microphone elements 102a, b, c, . . . , z may have other directionalities or polar patterns, such as cardioid, subcardioid, or omnidirectional.

Each of the microphone elements 102a, b, c, . . . , z in the array microphone may detect sound and convert the sound to an audio signal. Components in the array microphone, such as analog to digital converters, processors, and/or other components, may process the audio signals and ultimately generate one or more digital audio output signals. In other embodiments, the microphone elements 102a, b, c, . . . , z in the array microphone may output analog audio signals so that other components and devices (e.g., processors, mixers, recorders, amplifiers, etc.) external to the array microphone may process the analog audio signals.

In embodiments, an audio device may include the microphone elements 102a, b, c, . . . , z of the array microphone and also one or more loudspeakers. The loudspeakers may be utilized, for example, to play sound from a remote location of a teleconference and/or to play other sounds. The microphone elements and the loudspeaker may be located in the same housing, in some embodiments, or may be in separate housings, in other embodiments. The sound from the loudspeakers may be considered an undesired sound, and accordingly, the location of the loudspeaker may be considered an undesired sound location.

The coefficient generator 104 may determine the coefficients for the beamformer 106. The beamformer 106 may utilize the coefficients generated by the coefficient generator 104 to create a beamformed signal associated with one or more lobes steered towards desired sound locations and one or more nulls steered towards undesired sound locations. The coefficient generator 104 may determine the coefficients for the beamformer 106 based on the steering vectors for the locations of desired sound sources and undesired sound sources. In embodiments, desired widths and/or other parameters of lobes and/or nulls may also be utilized by the coefficient generator 104 to determine coefficients for the beamformer 106. The steering vectors for the locations of desired sound sources and undesired sound sources may be determined or configured as a particular three-dimensional coordinate relative to the location of the array microphone, such as in Cartesian coordinates (i.e., x, y, z), or in spherical coordinates (i.e., radial distance r, polar angle θ (theta), azimuthal angle φ (phi)), for example.

The coefficient generator 104 may generate the coefficients for the beamformer 106 such that the lobes associated with the resulting beamformed signal have a positive unity gain and such that the nulls associated with the resulting beamformed signal have an attenuated gain. In embodiments, the gains associated with the lobes may be less than a unity gain. In embodiments, the nulls may have a completely attenuated gain or a less than completely attenuated gain. The attenuated gain associated with the nulls may be less than completely attenuated, for example, in order to assist with maintaining the consistency of the shapes of the lobes of a pickup pattern. For example, a null of a pickup pattern may be configured to not be completely attenuated so that a lobe of the pickup pattern can still have a shape that can adequately detect and capture a desired sound source. This may be beneficial in situations where the steering vectors of a lobe and a null are relatively close together, and where having the null with a completely attenuated gain may negatively impact the pickup of desired sound sources by the lobe.

Optimized coefficients for the beamformer 106 may be generated by the coefficient generator 104 using a calculation process involving matrix inversion and/or Lagrange multipliers. Such optimized coefficients may result in lobes and nulls that are more precisely steered towards desired sound sources and undesired sound sources, respectively. The calculation process described herein may also be more efficient and result in quicker generation of the coefficients. An exemplary embodiment of the calculation process is described below with reference to FIG. 3 showing a method 206 for generating coefficients for the beamformer 106 using the coefficient generator 104.

In embodiments, the steering vectors for the locations of desired sound sources and undesired sound sources may be determined by an audio activity localizer or other suitable component that can determine the locations of audio activity in an environment based on the audio signals from the microphone elements 102a, b, c, . . . , z. For example, the audio activity localizer my utilize a Steered-Response Power Phase Transform (SRP-PHAT) algorithm, a Generalized Cross Correlation Phase Transform (GCC-PHAT) algorithm, a time of arrival (TOA)-based algorithm, a time difference of arrival (TDOA)-based algorithm, or another suitable sound source localization algorithm. In embodiments, the audio activity localizer may be included in the system 100, may be included in another component, or may be a standalone component.

In other embodiments, the steering vectors for the locations of desired sound sources and undesired sound sources may be determined programmatically or algorithmically using automated decision-making schemes, e.g., automatic focusing, placement, and/or deployment of a beam. Embodiments of such schemes are described in commonly assigned U.S. patent application Ser. Nos. 16/826,115 and 16/887,790, which are hereby incorporated by reference in their entirety herein. In further embodiments, the steering vectors for the locations of desired sound sources and undesired sound sources may be manually configured by a user, e.g., via a user interface on an electronic device in communication with the coefficient generator 104. In still other embodiments, the steering vectors for the locations of desired sound sources and undesired sound sources may be adaptively determined.

In embodiments, the null generation functionality of the coefficient generator 104 may be enabled or disabled, such as by a user. It may be desirable in some scenarios to disable the null generation functionality of the coefficient generator 104 for testing and other purposes, such as if it is desired that all of the audio sources in an environment are to be transmitted to a remote location during a teleconference. When the null generation functionality of the coefficient generator 104 is disabled, the coefficient generator 104 may determine coefficients for the beamformer 106, which the beamformer 106 may utilize to create a beamformed signal associated with one or more lobes steered towards desired sound locations. In this scenario, the coefficient generator 104 may determine coefficients for the beamformer 106 based on the steering vectors for the locations of the desired sound sources.

When the null generation functionality of the coefficient generator 104 is enabled, the coefficient generator 104 may determine the coefficients for the beamformer 106 based on the steering vectors for the locations of the desired sound sources and the steering vectors for the locations of the undesired sound sources. The beamformer 106 may utilize the coefficients to create a beamformed signal associated with one or more lobes steered towards desired sound locations and one or more nulls steered toward undesired sound sources.

Audio signals from the microphone elements 102a, b, c, . . . , z may be received at the beamformer 106. The beamformer 106 may generate one or more beamformed signals based on the audio signals from the microphone elements 102a, b, c, . . . , z and the coefficients from the coefficient generator 104. The beamformed signals generated by the beamformer 106 may be associated with one or more lobes and/or nulls of a desired pickup pattern. In embodiments, the beamformer 106 may utilize a frequency domain beamforming technique and/or a time domain beamforming technique.

In embodiments, the frequency domain beamforming technique utilized by the beamformer 106 may be a superdirective beamforming technique (such as a minimum variance distortionless response (MVDR) beamforming technique), a delay and sum beamforming technique, and/or another appropriate beamforming technique. In embodiments, the time domain beamforming technique utilized by the beamformer 106 may be a delay and sum beamforming technique, and/or another appropriate beamforming technique. Exemplary embodiments of hybrid audio beamforming systems and methods that may be implemented in the beamformer 106 are described in commonly assigned U.S. Provisional Pat. App. No. 63/142,711 and U.S. patent application Ser. No. 17/586,213, which are hereby incorporated by reference in their entirety herein.

The beamformer 106 may generate the one or more beamformed signals so that the associated null includes attenuation at one or more frequency ranges, such as a frequency range that typically includes human voice or any other frequency range. For example, the frequency range that is attenuated may be between 0 to 2.4 kHz, in some embodiments, or between 0 to 12 kHz in other embodiments. By attenuating certain frequency ranges, the beamformed signals generated by the beamformer 106 may minimize the inclusion of audio that would need to be processed (and possibly cancelled) by the acoustic echo canceller 110.

In embodiments, the frequency ranges that a lobe and/or null is attenuated at may be selected such that an echo return loss enhancement (ERLE) metric of the acoustic echo canceller 110 and/or an echo return loss (ERL) metric is increased. The ERLE metric is a measure of the performance of the acoustic echo canceller 110, and may indicate how much echo has been attenuated from an audio signal. The ERL metric is a ratio of the reference signal (e.g., far end remote signal) and the measured echo in the beamformed signal. In other embodiments, the frequency ranges that a lobe and/or null is attenuated at may be selected based on the ERLE metric of the acoustic echo canceller 110 and/or based on the ERL metric.

The mixer 108 may combine the beamformed signals from the beamformer 106 to generate a mixed beamformed signal. In embodiments, the mixer 108 may gate and/or attenuate a particular beamformed signal to mitigate the contribution of that beamformed signal in the mixed beamformed signal. The mixer 108 may combine the beamformed signals from the beamformer 106 with one or more audio signals from other sources, in some embodiments.

The acoustic echo canceller 110 may receive the mixed beamformed signal from the mixer 108, perform acoustic echo cancellation on the mixed beamformed signal, and generate an echo-cancelled mixed beamformed signal. The echo-cancelled mixed beamformed signal may include mitigation of the sound in a reference audio signal. The reference audio signal may include, for example, the sound received from a remote location and/or locally generated or played sounds that may be picked up by the array microphone and are desired to be removed from the mixed beamformed signal.

The echo-cancelled mixed beamformed signal from the acoustic echo canceller 110 may be transmitted to components or devices (e.g., processors, recorders, amplifiers, etc.) external to the system 100. In embodiments, the echo-cancelled mixed beamformed signal may be transmitted to a remote location (e.g., a far end of a teleconference) and/or played in the local environment for sound reinforcement. In other embodiments, the mixed beamformed signal from the mixer 108 may be transmitted to components or devices external to the system 100 and/or to a remote location, in addition to or in lieu of the echo-cancelled mixed beamformed signal from the acoustic echo canceller 110. In this way, the echo-cancelled mixed beamformed signal may be, for example, transmitted to a remote location without the undesirable echo of persons at the remote location hearing their own speech and sound.

In embodiments, the acoustic echo canceller 110 may process the mixed beamformed signal through a post-mix acoustic echo cancellation algorithm. In such embodiments, the acoustic echo canceller 110 may include a signal selection mechanism that is configured to select at least one of the beamformed signals such that the echo-cancelled mixed beamformed signal is generated based on the mixed beamformed signal, information gathered from the selected beamformed signal, and the reference audio signal. Information gathered from the selected beamformed signal may include, for example, measurements of the background error power and hidden error power of the selected beamformed signal. The signal selection mechanism may include a switch, a mixer that could select a particular beamformed signal (by attenuating some or all of the other beamformed signals), and/or another suitable signal selection mechanism. Exemplary embodiments of post-mix acoustic echo cancellation systems and methods are described in commonly-assigned U.S. Pat. No. 10,367,948 entitled “Post-Mixing Acoustic Echo Cancellation Systems and Methods”, which is incorporated by reference in its entirety herein. In some embodiments, the echo-cancelled mixed beamformed signal may be further processed to reduce noise.

An embodiment of a method 200 for beamforming with nulling control, mixing, and echo cancelling of audio signals using the system 100 is shown in FIG. 2. The method 200 may be utilized to generate an echo-cancelled mixed beamformed signal using the system 100 shown in FIG. 1, where the echo-cancelled mixed beamformed signal may be associated with lobes that are steered towards desired sound locations and also associated with nulls that are steered towards undesired sound locations. The beamformer coefficients that are generated in the method 200 may be the result of an optimized calculation process involving matrix inversion and Lagrange multipliers. One or more processors and/or other processing components (e.g., analog to digital converters, encryption chips, etc.) within or external to the system 100 may perform any, some, or all of the steps of the method 200. One or more other types of components (e.g., memory, input and/or output devices, transmitters, receivers, buffers, drivers, discrete components, etc.) may also be utilized in conjunction with the processors and/or other processing components to perform any, some, or all of the steps of the method 200.

At step 202, audio signals from the microphone elements 102a, b, c, . . . , z may be received at the beamformer 106. Steering vectors associated with the locations of one or more desired sound locations and/or one or more undesired sound locations may be received at step 204 at the coefficient generator 104. At step 206, the coefficient generator 104 may generate beamformer coefficients based on the steering vectors associated with the desired sound locations and the undesired sound locations received at step 204. The beamformer coefficients generated at step 206 may be utilized to generate beamformed signals associated with one or more lobes and/or one or more nulls. The lobes may be steered towards the desired sound locations and the nulls may be steered towards the undesired sound locations. In embodiments, an optimized set of beamformer coefficients may be generated at step 206, as described in more detail below with reference to FIG. 3. Steps 202, 204, and/or 206 may be performed substantially at the same time or may be performed at different times.

At step 208, the beamformer 106 may generate beamformed signals that are associated with one or more lobes and/or nulls of a desired pickup pattern, e.g., lobes steered towards desired sound locations and/or nulls steered towards undesired sound locations. The beamformed signals may be generated at step 208 based on the audio signals received from the microphone elements 102a, b, c, . . . , z at step 202 and based on the beamformer coefficients generated at step 206. The beamformer 106 may utilize a frequency domain beamforming technique and/or a time domain beamforming technique at step 208. In embodiments, the lobes associated with the beamformed signal may have a unity gain and the nulls associated with the beamformed signal may have an attenuated gain. In this way, the lobes can be steered to detect the audio of desired sound (e.g., talkers) and the nulls can be steered to ignore the audio of undesired sound (e.g., sound from loudspeakers).

At step 210, the mixer 108 may combine the beamformed signals generated by the beamformer at step 208 to create a mixed beamformed signal. In embodiments, the mixer 108 may generate the mixed beamformed signal at step 210 such that a desired audio mix is created where audio from certain beamformed signals is emphasized while audio from other beamformed signals is deemphasized or suppressed. At step 212, the acoustic echo canceller 110 may perform acoustic echo cancellation on the mixed beamformed signal created at step 210 by the mixer 108 to generate an echo-cancelled mixed beamformed signal.

FIG. 3 shows an embodiment of a method 206 for calculating beamformer coefficients based on steering vectors of desired sound sources and undesired sound sources. The method 206 shown in FIG. 3 may correspond to step 206 of the method 200 shown in FIG. 2. In the method 206 of FIG. 3, the audio signals from the microphone elements 102a, b, c, . . . , z may be utilized at step 302 to form a noise covariance matrix. The input audio signals x (n) from the microphone elements 102a, b, c, . . . , z may be represented as:

$\begin{array}{cc}x\left(n\right)\stackrel{△}{=}{x}_s\left(n\right)+x_{\xi }\left(n\right),& \left(1\right)\end{array}$

where x_s (n) is a column vector of the desired signals and x_ξ(n) is the column vector of the undesired signals (e.g., noise and/or other interference). As seen in equation (1), the input audio signals may be the sum of the desired signals and the undesired signals that are detected in the environment by the array microphone. The output y(n) of the beamformer at a time instant n may be represented as:

$\begin{array}{cc}y\left(n\right)\stackrel{△}{=}{W}^{H}x\left(n\right)={\sum }_{m=0}^{M-1}{w}_m^{*}{x}_m\left(n\right),& \left(2\right)\end{array}$

where W is a complex vector representing the beamformer coefficients and ( )^H is a Hermitian transpose. As seen in equation (2), the output y(n) of the beamformer may be based on the beamformer coefficients and the input audio signals.

The noise variance matrix may represent the amount of noise and/or other undesired sounds in an environment. The size of the noise variance matrix may be relatively large, such as based on the number of microphone elements 102a, b, c, . . . , z, e.g., M by M, where M is the number of microphone elements 102a, b, c, . . . , z. The noise covariance matrix R_ξ may be represented as:

$\begin{array}{cc}{R}_{\xi }\stackrel{△}{=}E\left[x_{\xi }\left(n\right){x_{\xi }\left(n\right)}^H\right],& \left(3\right)\end{array}$

where ( )^H represents a Hermitian transpose. As seen in equation (3), the noise covariance matrix may be created based on the undesired signals x_ξ(n).

At step 304, a null matrix may be formed based on the steering vectors of the undesired sound sources received at step 204. The null matrix may represent the locations of the desired sound sources in an environment (e.g., where lobes may be steered towards) and the locations of the undesired sound sources in an environment (e.g., where nulls may be steered towards), and may also include an inversion of the noise covariance matrix formed at step 302. The size of the null matrix may be relatively small, such as based on the number of nulls to be generated. In embodiments, the size of the null matrix may be (1+N) by (1+N), where N is the number of nulls. The null matrix Φ may be represented as:

$\begin{array}{cc}\Phi =\left[\begin{array}{ccc}{\Phi }_{00}{\Phi }_{01}& \dots & {\Phi }_{0L}\\ ⋮& \ddots & ⋮\\ {\Phi }_{L0}{\Phi }_{L1}& \dots & {\Phi }_{\mathrm{LL}}\end{array}\right],\mathrm{and}& \left(4\right)\end{array}$ ${\Phi }_{ij}={v\left({\theta }_i\right)}^H{R}_{\xi }^{-1}v\left({\theta }_j\right),i,j=0,\dots ,N,$

where v(θ_i) represents the steering vectors for each of the desired sound sources and the undesired sound sources. In particular, steering vector v(θ_i) may represent delays in the frequency domain which depend on the microphone constellation of the array and the direction of the desired sound sources and/or undesired sound sources. As seen in equation (4), the null matrix Φ may be based on the steering vectors for each of the desired sound sources and the undesired sound sources, and the inverted noise covariance matrix.

At step 306, the beamformer coefficients may be calculated based on the noise covariance matrix formed at step 302, the null matrix formed at step 304, and the steering vectors of desired sound sources and undesired sound sources received at step 204. The beamformer coefficients may be calculated at step 306 using a Lagrange multiplier that includes an inversion of the null matrix. The beamformer coefficients can be solved for by minimizing the variance of noise at the output of the beamformer 106 and minimizing the total power P from the beamformer 106, subject to a unity gain for lobes steered towards desired sound sources and an attenuated gain for the nulls steered towards undesired sound sources.

The minimization of the variance of noise can be represented as:_W^minW^HR_ξW,subject to W^Hv(θ_s)=1, and W^Hv(θ_ni)=p_i,for all i in [1,N]  (5),

where p_i is an attenuation value for the nulls and N is the number of nulls, and v(θ_s) and v (θ_ni) represent the steering vectors of desired sound sources and undesired sound sources, respectively. The minimization of the total power from the beamformer P can be represented as:P=W^HR_ξW+λ_s[W^Hv(θ_s)−1]+Σ_i=1^Nλ_i[W^Hv(θ_ni)−p_i]  (6),

where λ_s and λ_i are Lagrange multipliers and p_i is the amount of attenuation for the nulls that are being steered towards the undesired sound sources.

Finally, the optimum beamformer coefficients W_opt can be calculated as follows:

$\begin{array}{cc}{W}_{\mathrm{opt}}=-R_{\xi }^{-1}\left[{\sum }_{i=0}^{N}v\left({\theta }_i\right){\lambda }_i\right],& \left(7\right)\end{array}$

where v(θ_i)=v(θ_s) for i=0, and v(θ_i)=v (θ_ni) for i=1, . . . , N.

As seen in equation (7), the optimum beamformer coefficients may be based on the inverted noise covariance matrix, the steering vectors for each of the desired sound sources and the undesired sound sources, and the Lagrange multipliers.

The beamformer coefficients calculated at step 306 may be utilized by the beamformer 106 at step 208, as described previously. In embodiments, the beamformer coefficients may be calculated at step 306 to take into account the physical configuration of microphone elements 102a, b, c, . . . , z of the array microphone. For example, the microphone elements 102a, b, c, . . . , z may be arranged in a one-dimensional, two-dimensional, or three-dimensional configuration, and/or may be arranged in any suitable shape. As such, depending on the configuration of the microphone elements 102a, b, c, . . . , z, the beamformer coefficients may be calculated at step 306 when there is no conflict between the steering vectors of desired sound sources and undesired sound sources. For example, in a one-dimensional configuration of the microphone elements 102a, b, c, . . . , z, the beamformer coefficients may be calculated only if the azimuth for the steering vectors of desired sound sources and undesired sound sources are not the same.

FIGS. 4A-4D show exemplary diagrams depicting various scenarios of an array microphone in an environment and using the system 100 described above to more precisely control lobes and nulls.

In the exemplary scenario depicted in FIG. 4A, a pickup pattern of an array microphone 402 may include a single lobe 404 steered towards a desired sound source 406 and a single null 408 steered towards an undesired sound source 410. For example, as depicted in FIG. 4B, there may be a desired source 416 (e.g., a talker) in a location that a single lobe 414 of the array microphone 412 is steered towards to detect sound from, and there may be a loudspeaker 420 in physical proximity to the array microphone 412 that a single null 418 is steered towards to minimize detection of the sound from the loudspeaker 420. In embodiments, the array microphone 412 and the loudspeaker 420 may be in the same housing.

In another exemplary scenario, a pickup pattern of the array microphone may include a single lobe steered towards the desired sound source and multiple nulls steered towards one or more undesired sound sources. For example, as depicted in FIG. 4C, multiple nulls 428 may be steered towards the location of a loudspeaker 430 in order to maximize the attenuation of the sound from the loudspeaker 430, and a single lobe 424 of the array microphone 422 may be steered towards the desired sound source 426. In embodiments, the multiple nulls may be steered to cover an effective area of a relatively large loudspeaker.

In a further exemplary scenario, a pickup pattern of the array microphone may include multiple lobes steered towards one or more desired sound sources and one or more nulls steered towards one or more undesired sound sources. For example, as depicted in FIG. 4D, multiple lobes 434 of an array microphone 432 may be steered towards several desired sound sources 436 (e.g., active talkers) in an environment, while multiple nulls 438 may be steered towards multiple loudspeakers 440 in the environment.

Any process descriptions or blocks in figures should be understood as representing modules, segments, or portions of code which include one or more executable instructions for implementing specific logical functions or steps in the process, and alternate implementations are included within the scope of the embodiments of the invention in which functions may be executed out of order from that shown or discussed, including substantially concurrently or in reverse order, depending on the functionality involved, as would be understood by those having ordinary skill in the art.

This disclosure is intended to explain how to fashion and use various embodiments in accordance with the technology rather than to limit the true, intended, and fair scope and spirit thereof. The foregoing description is not intended to be exhaustive or to be limited to the precise forms disclosed. Modifications or variations are possible in light of the above teachings. The embodiment(s) were chosen and described to provide the best illustration of the principle of the described technology and its practical application, and to enable one of ordinary skill in the art to utilize the technology in various embodiments and with various modifications as are suited to the particular use contemplated. All such modifications and variations are within the scope of the embodiments as determined by the appended claims, as may be amended during the pendency of this application for patent, and all equivalents thereof, when interpreted in accordance with the breadth to which they are fairly, legally and equitably entitled.

Claims

1. A method, comprising: receiving a plurality of audio signals from a plurality of microphones;receiving a first steering vector associated with a desired sound location and a second steering vector associated with an undesired sound location;generating a set of beamformer coefficients based on the first steering vector and the second steering vector, wherein the set of beamformer coefficients is generated to result in a beamformed signal associated with a lobe steered towards the desired sound location and a null steered towards the undesired sound location; andgenerating the beamformed signal based on the plurality of audio signals and the set of beamformer coefficients, using a frequency domain beamforming technique.

2. The method of claim 1, wherein the undesired sound location comprises a location of a loudspeaker.

3. The method of claim 1, wherein generating the set of beamformer coefficients comprises generating the set of beamformer coefficients such that the lobe steered towards the desired sound location has a unity gain and the null steered towards the undesired sound location has an attenuated gain.

4. The method of claim 3, wherein the attenuated gain associated with the null is variable relative to a shape of the lobe.

5. The method of claim 1, wherein generating the set of beamformer coefficients comprises generating an optimized set of beamformer coefficients using a Lagrange multiplier.

6. The method of claim 5, wherein generating the optimized set of beamformer coefficients using the Lagrange multiplier comprises: forming a noise covariance matrix based on undesired sound in the plurality of audio signals;forming a null matrix based on the noise covariance matrix, the first steering vector associated with the desired sound location, and the second steering vector associated with the undesired sound location; andcalculating the optimized set of beamformer coefficients using the Lagrange multiplier and based on the noise covariance matrix, the null matrix, the first steering vector, and the second steering vector.

7. The method of claim 1, further comprising: mixing the beamformed signal with at least one other beamformed signal to generate a mixed beamformed signal; andperforming acoustic echo cancellation on the mixed beamformed signal to generate an echo-cancelled mixed beamformed signal.

8. The method of claim 1, wherein generating the beamformed signal comprises generating the beamformed signal such that the null steered towards the undesired sound location is limited to attenuate a specific frequency range.

9. The method of claim 8, wherein the null steered towards the undesired sound location is limited to attenuate the specific frequency range such that an echo return loss metric of an acoustic echo cancellation process is increased.

10. The method of claim 8, further comprising selecting the specific frequency range based on an echo return loss metric of an acoustic echo cancellation process.

11. The method of claim 1, wherein the set of beamformer coefficients is generated to result in the beamformed signal being further associated with (1) a plurality of lobes each steered towards the desired sound location and comprising the lobe, and (2) a plurality of nulls each steered towards the undesired sound location and comprising the null.

12. The method of claim 1, wherein the set of beamformer coefficients is generated to result in the beamformed signal being further associated with (1) the lobe steered towards the desired sound location, and (2) a plurality of nulls each steered towards a plurality of undesired sound locations and comprising the null.

13. The method of claim 1, wherein the set of beamformer coefficients is generated to result in the beamformed signal being further associated with (1) a plurality of lobes steered towards the desired sound location and comprising the lobe, and (2) the null steered towards the undesired sound location.

14. An audio device, comprising: a plurality of microphones configured to generate a plurality of audio signals;a coefficient generator configured to: receive a first steering vector associated with a desired sound location and a second steering vector associated with an undesired sound location; andgenerate a set of beamformer coefficients based on the first steering vector and the second steering vector, wherein the set of beamformer coefficients is generated to result in a beamformed signal associated with a lobe steered towards the desired sound location and a null steered towards the undesired sound location; anda beamformer in communication with the plurality of microphones and the coefficient generator, the beamformer configured to generate the beamformed signal based on the plurality of audio signals and the set of beamformer coefficients, using a frequency domain beamforming technique.

15. The audio device of claim 14: further comprising a loudspeaker; andwherein the undesired sound location comprises a location of the loudspeaker.

16. The audio device of claim 14, wherein the coefficient generator is configured to generate the set of beamformer coefficients by generating an optimized set of beamformer coefficients using a Lagrange multiplier.

17. The audio device of claim 14, further comprising: a mixer in communication with the beamformer, the mixer configured to mix the beamformed signal with at least one other beamformed signal to generate a mixed beamformed signal; andan acoustic echo canceller in communication with the mixer, the acoustic echo canceller configured to perform acoustic echo cancellation on the mixed beamformed signal to generate an echo-cancelled mixed beamformed signal.

18. The audio device of claim 14, wherein the beamformer is configured to generate the beamformed signal by generating the beamformed signal such that the null steered towards the undesired sound location is limited to attenuate a specific frequency range.

19. The audio device of claim 18, wherein the null steered towards the undesired sound location is limited to attenuate the specific frequency range such that an echo return loss metric of an acoustic echo cancellation process is increased.

20. The audio device of claim 18, wherein the beamformer is further configured to select the specific frequency range based on an echo return loss metric of an acoustic echo cancellation process.