Abstract: Embodiments include a processing device communicatively coupled to a plurality of audio devices comprising at least one microphone and at least one speaker, and to a digital signal processing (DSP) component having a plurality of audio input channels for receiving audio signals captured by the at least one microphone, the processing device being configured to identify one or more of the audio devices based on a unique identifier associated with each of the one or more audio devices; obtain device information from each identified audio device; and adjust one or more settings of the DSP component based on the device information. A computer-implemented method of automatically configuring an audio conferencing system, comprising a digital signal processing (DSP) component and a plurality of audio devices including at least one speaker and at least one microphone, is also provided.

Abstract: Audio-visual systems and methods are configured to determine a first talker location based on a first group of sound locations corresponding to audio detected by the microphone in association with one or more talkers; receive a new sound location for new audio detected by the microphone in association with at least one talker; determine a proximity of the new sound location to the first group of sound locations; based on the new sound location being in close proximity to one or more of the sound locations in the first group, determine a second talker location based on the new sound location and the first group of sound locations; determine a second proximity of the second talker location to the first talker location; provide the second talker location to the camera if the second proximity meets or exceeds a threshold; and otherwise, provide the first talker location the camera.

Abstract: Conferencing systems and methods configured to generate talker coordinates for directing a camera towards talker locations in an environment are disclosed, as well as talker tracking using multiple microphones and multiple cameras. One method includes determining, using a first microphone array and based on audio associated with a talker, a first talker location in a first coordinate system relative to the first microphone array; determining, using a second microphone array and based on the audio associated with the talker, a second talker location in a second coordinate system relative to the second microphone array; determining, based on the first talker location and the second talker location, an estimated talker location in a third coordinate system relative to a camera; and transmitting, to the camera, the estimated talker location in the third coordinate system to cause the camera to point the camera towards the estimated talker location.

Abstract: Various embodiments of the present disclosure provide methods, apparatus, systems, devices, and/or the like for inferring characteristics of a physical enclosure using a plurality of audio signals. The plurality of audio signals may be processed using a feature extraction framework to generate structured audio event data sets, which may be processed using an audio event framework to determine the characteristics of the physical enclosure.

Abstract: Embodiments include a method of reducing echo in an audio system comprising a microphone, an acoustic echo canceller (AEC), and at least one processor, the method comprising receiving, by the at least one processor, an audio signal detected by the microphone; deploying, by the at least one processor, a microphone lobe towards a first location associated with the detected audio signal; obtaining, by the at least one processor, one or more AEC parameters for the first location, the one or more AEC parameters being stored in a memory in communication with the at least one processor; initializing, by the at least one processor, the AEC using the one or more AEC parameters; and generating, by the at least one processor, an echo-cancelled output signal using the initialized AEC and based on the detected audio signal and a reference signal provided to the AEC.

Abstract: Embodiments include an audio system comprising a plurality of microphones disposed in an environment, wherein the plurality of microphones is configured to detect one or more audio sources, and generate location data indicating a location of each of the one or more audio sources relative to the plurality of microphones; and at least one processor communicatively coupled to the plurality of microphones, wherein the at least one processor is configured to receive the location data from the plurality of microphones, and define a plurality of audio pick-up regions in the environment based on the location data, the plurality of audio pick-up regions comprising a first audio pick-up region and a second audio pick-up region, wherein the plurality of microphones are configured to deploy a first lobe within the first audio pick-up region and a second lobe within the second audio pick-up region.

Abstract: Techniques for adaptively providing acoustic echo cancellation (AEC) for a stereo audio signal associated with at least one microphone are discussed herein. Some embodiments may include determining, based at least in part on detecting a reference signal associated with a channel sample portion of the stereo audio signal, a panning state of the stereo audio signal. A hard-panned-configured AEC processing filter or a soft-panned-configured AEC processing filter is applied to the stereo audio signal to generate a filtered audio signal output based on the panning state.

Abstract: Conferencing systems and methods configured to generate true talker coordinates for use in camera tracking of talkers and objects in an environment and other room intelligence use cases are disclosed. The initial configuration and ongoing usage of conferencing systems can be improved by detecting and converting the locations of objects and talkers in an environment into a common coordinate system. The amount of time and effort by installers, integrators, and users, can be reduced leading to increased satisfaction with installation and usage of the conferencing system.

Abstract: Embodiments include a processing device communicatively coupled to a plurality of audio devices comprising at least one microphone and at least one speaker, and to a digital signal processing (DSP) component having a plurality of audio input channels for receiving audio signals captured by the at least one microphone, the processing device being configured to identify one or more of the audio devices based on a unique identifier associated with each of said one or more audio devices; obtain device information from each identified audio device; and adjust one or more settings of the DSP component based on the device information. A computer-implemented method of automatically configuring an audio conferencing system, comprising a digital signal processing (DSP) component and a plurality of audio devices including at least one speaker and at least one microphone, is also provided.

Abstract: When the noise in an audio signal made up of both speech and noise is suppressed, the quality of the speech in the audio signal is usually degraded. The speech obtained from a noise-suppressed audio signal is improved by determining linear predictive coding (LPC) characteristics of the audio signal without or prior to noise suppression and by determining the LPC characteristics of the noise-suppressed audio. The convolution of those differing characteristics provides an improved-quality speech signal, with the original noise level reduced or suppressed.

Abstract: Operating parameters of a hands-free audio system used with a wireless communication device in a moving vehicle are adjusted or tuned in real-time and requires only two persons: one to drive the vehicle and thus provide actual usage conditions with the hands-free audio system and one to remotely tune or adjust operating parameters to optimize far end audio quality. The system is remotely tuned by transmitting audio signals from the vehicle to the far end using a first communications link to the far end and sending adjustment commands to the vehicle from the far end via a second, data link between the far end and the vehicle. In one embodiment, DTMF signals received from inside or outside the vehicle can tune or be used to diagnose the hands-free system. Test measurements obtained from within and by the hands-free audio system can also be retrieved from a remote location.

Abstract: Memory is dynamically shared or allocated in an embedded computer system. The types of memory that are part of the system are first determined. Thereafter, the amount of memory available for use is determined. The type of memory required by a program or application is determined as is the amount of space that is required. If the amount of memory space that can be allocated to the program in a first type of requested memory is greater than or at least equal to the amount of memory space required by the computer program, the program is then loaded into the available memory. If the requested type of memory is not available or there is not enough of the requested memory available, other types of memory devices are considered and used, if sufficient space in one or more of them exists.

Abstract: Speech in a motor vehicle is improved by suppressing transient, “non-stationary” noise using pattern matching. Pre-stored sets of linear predictive coefficients are compared to LPC coefficients of a noise signal. The pre-stored LPC coefficient set that is “closest” to an LPC coefficient set representing a signal comprising speech and noise is considered to be noise.

Abstract: A maximum noise suppression level (Gmin) is not a single constant value for an entire frequency range, but is allowed to vary across frequencies. The amount of variation is dynamically computed based on the input noise characteristics. For example, if there is excess noise in the lower frequency region, the maximum noise suppression level in that region will increase to suppress the noise in that frequency region. This feature can be enabled all the time, and will be active when the input conditions warrant extra noise suppression in a particular frequency region. Thus, the effort involved in manually tuning an audio system (e.g., hands-free telephony, voice-controlled automotive head unit, etc.) can be significantly reduced or eliminated.

Abstract: Speech in a motor vehicle is improved by suppressing transient, “non-stationary” noise using pattern matching. Pre-stored sets of linear predictive coefficients are compared to LPC coefficients of a noise signal. The pre-stored LPC coefficient set that is “closest” to an LPC coefficient set representing a signal comprising speech and noise is considered to be noise.

Abstract: A maximum noise suppression level (Gmin) is not a single constant value for an entire frequency range, but is allowed to vary across frequencies. The amount of variation is dynamically computed based on the input noise characteristics. For example, if there is excess noise in the lower frequency region, the maximum noise suppression level in that region will increase to suppress the noise in that frequency region. This feature can be enabled all the time, and will be active when the input conditions warrant extra noise suppression in a particular frequency region. Thus, the effort involved in manually tuning an audio system (e.g., hands-free telephony, voice-controlled automotive head unit, etc.) can be significantly reduced or eliminated.

Abstract: Acoustic noise in an audio signal is reduced by calculating a speech probability presence (SPP) factor using minimum mean square error (MMSE). The SPP factor, which has a value typically ranging between zero and one, is modified or warped responsive to a value obtained from the evaluation of a sigmoid function, the shape of which is determined by a signal-to-noise ratio (SNR), which is obtained by an evaluation of the signal energy and noise energy output from a microphone over time. The shape and aggressiveness of the sigmoid function is determined using an extrinsically-determined SNR, not determined by the MMSE determination. The extrinsically-determined SNR is obtained from a long term history of previously-determined speech presence probabilities and a long term history of previously-determined noise histories.

Abstract: Acoustic noise in an audio signal is reduced by calculating a speech probability presence (SPP) factor using minimum mean square error (MMSE). The SPP factor, which has a value typically ranging between zero and one, is modified or warped responsive to a value obtained from the evaluation of a sigmoid function, the shape of which is determined by a signal-to-noise ratio (SNR), which is obtained by an evaluation of the signal energy and noise energy output from a microphone over time. The shape and aggressiveness of the sigmoid function is determined using an extrinsically-determined SNR, not determined by the MMSE determination. The extrinsically-determined SNR is obtained from a long term history of previously-determined speech presence probabilities and a long term history of previously-determined noise histories.

Abstract: Differing first and second audio signal sample rates from first and second audio signals are matched to each other. If signal sample rates are different, a frame of samples of the first audio signal is duplicated. The duplicate copies are multiplied by a window function and its inverse to produce “windowed frames” first and last samples of which can be deleted or added to increase or decrease a frame rate.

Abstract: Acoustic noise in an audio signal is reduced by calculating a speech probability presence (SPP) factor using minimum mean square error (MMSE). The SPP factor, which has a value typically ranging between zero and one, is modified or warped responsive to a value obtained from the evaluation of a sigmoid function, the shape of which is determined by a signal-to-noise ratio (SNR), which is obtained by an evaluation of the signal energy and noise energy output from a microphone over time. The shape and aggressiveness of the sigmoid function is determined using an extrinsically-determined SNR, not determined by the MMSE determination. The extrinsically-determined SNR is obtained from a long term history of previously-determined speech presence probabilities and a long term history of previously-determined noise histories.