A system configured to perform deep adaptive acoustic echo cancellation (AEC) to improve audio processing. Due to mechanical noise and continuous echo path changes caused by movement of a device, echo signals are nonlinear and time-varying and not fully canceled by linear AEC processing alone. To improve echo cancellation, deep adaptive AEC processing integrates a deep neural network (DNN) and linear adaptive filtering to perform echo and/or noise removal. The DNN is configured to generate a nonlinear reference signal and step-size data, which the linear adaptive filtering uses to generate output audio data representing local speech. The DNN may generate the nonlinear reference signal by generating mask data that is applied to a microphone signal, such that the reference signal corresponds to a portion of the microphone signal that does not include near-end speech.

The present invention relates to an active noise cancellation integrated circuit for stacking multiple anti-noise signals, an associated method, and an active noise cancellation headphone using the same. The method is applicable to an audio playback device with multiple ANC filtering units. The method includes: acquiring an anti-noise signal from an ANC filtering unit; generating a decoupled signal by processing the anti-noise signal with the transfer function of a physical channel and operations of other ANC filtering units; performing a signal superposition, wherein an anti-noise signal from another ANC filtering unit is superposed with the decoupled signal; and performing an audio playback based on the superposed signal and an audio signal such that noise is eliminated.

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.

A voice communication technique by which superior echo canceling can be realized even in use of non-directional microphones is provided. When voice uttered by a near end speaker is defined as speaker's voice; sound obtained by emitting a received speech signal, which is a voice signal of a far end speaker, by a loudspeaker is defined as reproduction sound; a signal obtained by picking up an acoustic signal, which contains the speaker's voice and the reproduction sound, by a first microphone is defined as a first sound pickup signal; and a signal obtained by picking up an acoustic signal, which contains the speaker's voice and the reproduction sound, by a second microphone is defined as a second sound pickup signal, a voice communication device includes: a first subtraction unit that generates a first-subtracted first sound pickup signal that is a difference between a corrected sound pickup signal and the second sound pickup signal or the first sound pickup signal, the corrected sound pickup signal being obtained by correcting either one of the first sound pickup signal and the second sound pickup signal on the basis of an amplitude difference based on a difference between a distance d1 from the loudspeaker to the first microphone and a distance d2 from the loudspeaker to the second microphone (d2≤d1); and a second subtraction unit that generates a transmitted speech signal, which is to be transmitted to the far end speaker, by subtracting an adaptive-filtered second sound pickup signal, which is obtained by canceling echo, from the first-subtracted first sound pickup signal.

Systems and methods for audio listening devices, comprise a speaker coupled to a first housing, a sound port having a first end and a second end, wherein the first end is coupled to the first housing, and the second end is configured to be inserted in an ear canal of a person such that sound waves emitted from the speaker propagates via a secondary path to the ear canal through the sound port, active noise cancellation (ANC) components configured to generate anti-noise signals through the micro-speakers to cancel external noise, and a first microphone disposed within the sound port at the second end of the sound port such that the first microphone is configured to detect the anti-noise signal that propagates through the sound port via the secondary path and the external noise that propagates via a primary path.

A method for hybrid noise suppression includes receiving a processed audio signal from an audio device. The processed audio signal results from a partial audio processing performed on a noisy audio input signal. The method further includes predicting a noise suppression parameter using a neural network model operating on the processed audio signal and generating a noise-suppressed audio signal from the processed audio signal, using the noise suppression parameter. The method further includes generating a noise-suppressed audio output signal from the noise-suppressed audio signal using an additional audio processing and outputting the noise-suppressed audio output signal.

In some implementations, an output of a first channel of an echo cancellation variable filter having an output of a first microphone output from a second speaker as an input is added to an output of a second microphone. An echo cancellation coefficient updating unit updates the filter coefficient of the first channel so that the error that is the output of a second adder is minimized. Using the output of the second channel that uses the output of a sound source device output from the first speaker as an input and shares the filter coefficient with the first channel as a reference signal, and the output of the second microphone as an error, the noise cancellation coefficient updating unit updates the filter coefficient of the noise cancellation variable filter that generates a noise-canceling sound to be output from the output of the sound source device to the second speaker.

Embodiments include an audio system comprising a speaker array comprising a plurality of drivers arranged in a first planar configuration; a microphone array comprising a plurality of transducers arranged in a second planar configuration; and a beamformer communicatively coupled to the speaker array and the microphone array, the beamformer configured to: generate an individual speaker output signal for each of the drivers in the speaker array based on one or more input audio signals received from an audio source, and generate a microphone output signal for the microphone array based on one or more microphone signals captured by one or more of the transducers. Embodiments also include a method performed by one or more processors coupled to a microphone array having a plurality of transducers and a speaker array having a plurality of drivers.

A playback device is configured to: produce a first channel audio output of a first channel of audio content; produce a second channel audio output of a second channel of the audio content; receive captured audio content comprising (i) a first portion corresponding to the first channel audio output, (ii) a second portion corresponding to the second channel audio output, and (iii) a third portion corresponding to a voice command, wherein the captured audio content has a first signal-to-noise ratio; determine a set of signal components from at least one of the first channel or the second channel of the audio content; perform acoustic echo cancellation on a subset of signal components; determine an acoustic echo cancellation output; and apply the acoustic echo cancellation output to the captured audio content and thereby increase the first signal-to-noise ratio to a second signal-to-noise ratio that is greater than the first signal-to-noise ratio.

Disclosed are techniques for performing ray-based acoustic modeling that models scattering of acoustic waves by a surface of a device. The acoustic modeling uses two parameters, a room response representing acoustics and geometry of a room and a device response representing acoustics and geometry of the device. The room response is determined using ray-based acoustic modeling, such as ray tracing. The device response can be measured in an actual environment or simulated and represents an acoustic response of the device to individual acoustic plane waves. The device applies a superposition of the room response and the plane wave scattering from the device response to determine acoustic pressure values and generate microphone audio data. The device can estimate room impulse response (RIR) data using data from the microphones, and can use the RIR data to perform audio processing such as sound equalization, acoustic echo cancellation, audio beamforming, and/or the like.