Abstract: A first device obtains, from the array, several audio signals and processes the audio signals to produce a speech signal and one or more ambient signals. The first device processes the ambient signals to produce a sound-object sonic descriptor that has metadata describing a sound object within an acoustic environment. The first device transmits, over a communication data link, the speech signal and the descriptor to a second electronic device that is configured to spatially reproduce the sound object using the descriptor mixed with the speech signal, to produce several mixed signals to drive several speakers.

Abstract: A first device obtains, from the array, several audio signals and processes the audio signals to produce a speech signal and one or more ambient signals. The first device processes the ambient signals to produce a sound-object sonic descriptor that has metadata describing a sound object within an acoustic environment. The first device transmits, over a communication data link, the speech signal and the descriptor to a second electronic device that is configured to spatially reproduce the sound object using the descriptor mixed with the speech signal, to produce several mixed signals to drive several speakers.

Abstract: A first device obtains, from the array, several audio signals and processes the audio signals to produce a speech signal and one or more ambient signals. The first device processes the ambient signals to produce a sound-object sonic descriptor that has metadata describing a sound object within an acoustic environment. The first device transmits, over a communication data link, the speech signal and the descriptor to a second electronic device that is configured to spatially reproduce the sound object using the descriptor mixed with the speech signal, to produce several mixed signals to drive several speakers.

Abstract: A first device obtains, from the array, several audio signals and processes the audio signals to produce a speech signal and one or more ambient signals. The first device processes the ambient signals to produce a sound-object sonic descriptor that has metadata describing a sound object within an acoustic environment. The first device transmits, over a communication data link, the speech signal and the descriptor to a second electronic device that is configured to spatially reproduce the sound object using the descriptor mixed with the speech signal, to produce several mixed signals to drive several speakers.

Abstract: Provided are systems and methods for suppressing key clicks in audio signals. An example method includes extracting features of an audio signal. The features are provided as inputs to a neural network. The neural network is trained to identify clicks in the audio signal and/or generate a multiplicative suppression mask suitable for removing key clicks from the audio signal. The suppression mask is applied to the audio signal to produce a clicks-removed audio signal. Comfort noise may be added to the clicks-removed audio signal to avoid noise pumping artifacts. The example method can be used without imposing keyboard activity restrictions on users. The key click suppression can be used in audio systems with a single microphone or with multiple microphones.

Abstract: A method for restoring distorted speech components of an audio signal distorted by a noise reduction or a noise cancellation includes determining distorted frequency regions and undistorted frequency regions in the audio signal. The distorted frequency regions include regions of the audio signal in which a speech distortion is present. Iterations are performed using a model to refine predictions of the audio signal at distorted frequency regions. The model is configured to modify the audio signal and may include deep neural network trained using spectral envelopes of clean or undamaged audio signals. Before each iteration, the audio signal at the undistorted frequency regions is restored to values of the audio signal prior to the first iteration; while the audio signal at distorted frequency regions is refined starting from zero at the first iteration. Iterations are ended when discrepancies of audio signal at undistorted frequency regions meet pre-defined criteria.

Abstract: Systems and methods for improving performance of a directional audio capture system are provided. An example method includes correlating phase plots of at least two audio inputs, with the audio inputs being captured by at least two microphones. The method can further include generating, based on the correlation, estimates of salience at different directional angles to localize a direction of a source of sound. The method can allow providing cues to the directional audio capture system based on the estimates. The cues include attenuation levels. A rate of change of the levels of attenuation is controlled by attack and release time constants to avoid sound artifacts. The method also includes determining a mode based on an absence or presence of one or more peaks in the estimates of salience. The method also provides for configuring the directional audio capture system based on the determined mode.

Abstract: Provided are systems and methods for generating clean speech from a speech signal representing a mixture of a noise and speech. The clean speech may be generated from synthetic speech parameters. The synthetic speech parameters are derived based on the speech signal components and a model of speech using auditory and speech production principles. The modeling may utilize a source-filter structure of the speech signal. One or more spectral analyzes on the speech signal are performed to generate spectral representations. The feature data is derived based on a spectral representation. The features corresponding to the target speech according to a model of speech are grouped and separated from the feature data. The synthetic speech parameters, including spectral envelope, pitch data and voice classification data are generated based on features corresponding to the target speech.

Abstract: Systems and methods for estimating and tracking multiple attributes of multiple objects from multi-sensor data are provided. An exemplary method includes identifying features associated with sensor data. The sensor data represents data captured by at least one of a plurality of acoustic and non-acoustic sensors. Identification of the features associated with the sensor data may be based variously on detected sounds, motions, images, and the like. The exemplary method further includes determining, in parallel, multiple probable objects based at least in part on the identified features. Various embodiments of the method also include forming hypotheses based at least in part on associating identified features with the multiple probable objects and attributing the formed hypotheses to channels. Sequence of the formed hypotheses are constructed.

Abstract: The present technology provides adaptive noise reduction of an acoustic signal using a sophisticated level of control to balance the tradeoff between speech loss distortion and noise reduction. The energy level of a noise component in a sub-band signal of the acoustic signal is reduced based on an estimated signal-to-noise ratio of the sub-band signal, and further on an estimated threshold level of speech distortion in the sub-band signal. In various embodiments, the energy level of the noise component in the sub-band signal may be reduced to no less than a residual noise target level. Such a target level may be defined as a level at which the noise component ceases to be perceptible.

Abstract: A robust noise reduction system may concurrently reduce noise and echo components in an acoustic signal while limiting the level of speech distortion. The system may receive acoustic signals from two or more microphones in a close-talk, hand-held or other configuration. The received acoustic signals are transformed to frequency domain sub-band signals and echo and noise components may be subtracted from the sub-band signals. Features in the acoustic sub-band signals are identified and used to generate a multiplicative mask. The multiplicative mask is applied to the noise subtracted sub-band signals and the sub-band signals are reconstructed in the time domain.

Abstract: A system utilizing two pairs of microphones for noise suppression. Primary and secondary microphones may be positioned closely spaced to each other to provide acoustic signals used to achieve noise cancellation/suppression. An additional, tertiary microphone may be spaced with respect to either the primary microphone or the secondary microphone in a spread-microphone configuration for deriving level cues from audio signals provided by the tertiary and the primary or secondary microphone. The level cues are expressed via a level difference used to determine one or more cluster tracking control signal(s). The level difference-based cluster tracking signals are used to control adaptation of noise suppression. A noise cancelled primary acoustic signal and level difference-based cluster tracking control signals are used during post filtering to adaptively generate a mask to be applied to a speech estimate signal.

Abstract: The present technology provides a robust noise suppression system that may concurrently reduce noise and echo components in an acoustic signal while limiting the level of speech distortion. A time-domain acoustic signal may be received and be transformed to frequency-domain sub-band signals. Features, such as pitch, may be identified and tracked within the sub-band signals. Initial speech and noise models may be then be estimated at least in part from a probability analysis based on the tracked pitch sources. Speech and noise models may be resolved from the initial speech and noise models and noise reduction may be performed on the sub-band signals. An acoustic signal may be reconstructed from the noise-reduced sub-band signals.

Abstract: Systems and methods for improving performance of a directional audio capture system are provided. An example method includes correlating phase plots of at least two audio inputs, with the audio inputs being captured by at least two microphones. The method can further include generating, based on the correlation, estimates of salience at different directional angles to localize a direction of a source of sound. The method can allow providing cues to the directional audio capture system based on the estimates. The cues include attenuation levels. A rate of change of the levels of attenuation is controlled by attack and release time constants to avoid sound artifacts. The method also includes determining a mode based on an absence or presence of one or more peaks in the estimates of salience. The method also provides for configuring the directional audio capture system based on the determined mode.

Abstract: A method for restoring distorted speech components of an audio signal distorted by a noise reduction or a noise cancellation includes determining distorted frequency regions and undistorted frequency regions in the audio signal. The distorted frequency regions include regions of the audio signal in which a speech distortion is present. Iterations are performed using a model to refine predictions of the audio signal at distorted frequency regions. The model is configured to modify the audio signal and may include deep neural network trained using spectral envelopes of clean or undamaged audio signals. Before each iteration, the audio signal at the undistorted frequency regions is restored to values of the audio signal prior to the first iteration; while the audio signal at distorted frequency regions is refined starting from zero at the first iteration. Iterations are ended when discrepancies of audio signal at undistorted frequency regions meet pre-defined criteria.

Abstract: Systems and methods for utilizing level differences to attenuate noise and enhance speech are provided. In exemplary embodiments, energy estimates of acoustic signals, representing captured sound, are determined in order to determine a level difference. This level difference in combination with a noise estimate, based only on a primary acoustic signal, allow a filter estimate to be derived. In some embodiments, the derived filter estimate may be smoothed. The filter estimate is then applied to the acoustic signal from the primary acoustic signal to generate a speech estimate.

Abstract: The present technology provides adaptive noise reduction of an acoustic signal using a sophisticated level of control to balance the tradeoff between speech loss distortion and noise reduction. The energy level of a noise component in a sub-band signal of the acoustic signal is reduced based on an estimated signal-to-noise ratio of the sub-band signal, and further on an estimated threshold level of speech distortion in the sub-band signal. In various embodiments, the energy level of the noise component in the sub-band signal may be reduced to no less than a residual noise target level. Such a target level may be defined as a level at which the noise component ceases to be perceptible.

Abstract: Systems and methods for estimating and tracking multiple attributes of multiple objects from multi-sensor data are provided. An exemplary method includes identifying features associated with sensor data. The sensor data represents data captured by at least one of a plurality of acoustic and non-acoustic sensors. Identification of the features associated with the sensor data may be based variously on detected sounds, motions, images, and the like. The exemplary method further includes determining, in parallel, multiple probable objects based at least in part on the identified features. Various embodiments of the method also include forming hypotheses based at least in part on associating identified features with the multiple probable objects and attributing the formed hypotheses to channels. Sequence of the formed hypotheses are constructed.

Abstract: The present technology provides robust, high quality expansion of the speech within a narrow bandwidth acoustic signal which can overcome or substantially alleviate problems associated with expanding the bandwidth of the noise within the acoustic signal. The present technology carries out a multi-faceted analysis to accurately identify noise within the narrow bandwidth acoustic signal. Noise classification information regarding the noise within the narrow bandwidth acoustic signal is used to determine whether to expand the bandwidth of the narrow bandwidth acoustic signal. By expanding the bandwidth based on the noise classification information, the present technology can expand the speech bandwidth of the narrow bandwidth acoustic signal and prevent or limit the bandwidth expansion of the noise.

Abstract: An audio equivalent of a video zoom feature for video recording and communication applications, as well as video post production processes. The audio zoom may operate in conjunction with a video zoom feature or independently. The audio zoom may be achieved by controlling reverberation effects of a signal, controlling a gain of the signal, as well as controlling the width of a directional beam which is used to select the particular audio component to focus on. The audio zoom may operate in response to user input, such as a user selection of a particular direction, or automatically based a current environment or other factors.