Provided in the present disclosure are a voice processing method, an apparatus, an electronic device, and a storage medium, the method comprising: detecting the working state of a current call system, and when the working state is a two-end speaking state or a remote-end speaking state, performing compression processing on a subsequent remote-end voice signal, acquiring a near-end voice signal by means of a microphone, performing echo processing on the basis of the near-end voice signal and the compression-processed remote-end voice signal to obtain an echo-processed near-end voice signal and a remaining echo signal, performing non-linear suppression processing on the near-end voice signal and the remaining echo signal, and performing gain control on the suppression-processed near-end voice signal.

Systems, methods, and machine-readable storage devices that receive an input signal representing audio captured using a microphone. The input signal includes portions that represent acoustic output from one or more audio sources, and a portion that represents other acoustic energy in the environment. A frequency domain representation of the input signal is iteratively modified to substantially reduce effects due to all but a selected one of the portions, from which an estimate of the power spectral density, PSD, of the selected portion is determined. Based upon the estimated PSD a noise or echo component is reduced, or a replacement noise is provided. The iterative modification involves a diagonalization of the cross-spectral density matrix to remove content coherent with a first audio input from the auto and cross-spectra of other signals.

A method comprises: detecting noise in an environment with a microphone to produce a noise signal; receiving a voice signal to be played into the environment through a loudspeaker; performing multiband correction of the noise signal based on a microphone transfer function of the microphone, to produce a corrected noise signal; performing multiband correction of the voice signal based on a loudspeaker transfer function of the loudspeaker to produce a corrected voice signal; and computing multiband voice intelligibility results based on the corrected noise signal and the corrected voice signal.

System and method for detecting and identifying noises in a sound signal occurring during a call on a mobile device and selectively filtering and suppressing the noises in the sound signal are provided. In the mobile device, a processor is configured to receive a sound signal, detect noises in the received sound signal, identify the noises in the received sound signal, display the identified noises in a user interface (UI), receive a selection of the displayed identified noises from the UI, and filter the received selection of the displayed identified noises from the received sound signal. The processor may use a machine learning module with a neural network to detect and identify the noises in the received sound signal.

A method of noise reduction includes using a neural network to control a Wiener filter. The gains estimated by the neural network are combined with the gains produced by the Wiener filter. In this manner, the noise reduction system provides improved results as compared to using only a neural network.

An echo cancellation method based on delay estimation is provided. In the method, a microphone signal and a reference signal are received and preprocessed. In the preprocessed microphone signal and the preprocessed reference signal, frequency point signals with non-linearity in a current echo cancellation scenario are determined. A current delay estimation value is calculated based on frequency point signals without non-linearity in the microphone signal and the reference signal. The reference signal is shifted based on the current delay estimation value. An adaptive filter is updated based on the preprocessed microphone signal and the shifted reference signal, to perform echo cancellation.

A method for automatically detecting audio feedback in an input audio signal includes separately filtering the audio input signal with a plurality of separate analysis audio filters to generate a plurality of filtered audio signals. The separate analysis audio filters are different. Then, comparing at least two of the filtered audio signals to obtain an energy level difference. Performing one or more repetitions of the steps of filtering and comparing to establish a plurality of the energy level differences. Then comparing energy level differences from at least two of the repetitions to detect the audio feedback. The method includes features of automatically performing audio feedback suppression of the detected audio feedback.

A system and method for processing an audio input signal includes a microphone, a controller, and a communication link that may be coupled to a remote speaker. The microphone captures the audio input signal and communicates the audio input signal to the controller, and the controller is coupled to the communication link. The controller includes executable code to generate, via a linear noise reduction filtering algorithm, a first resultant based upon the audio input signal, and generate, via non-linear post filtering algorithm, a second resultant based upon the first resultant. An audio output signal is generated based upon the second resultant employing a feature restoration algorithm. The audio output signal is communicated, via the communication link, to a speaker that may be at a remote location.

A noise estimation method including steps of generating gap confidence values in response to microphone output and playback signals, and using the gap confidence values to generate an estimate of background noise in a playback environment. Each gap confidence value is indicative of confidence of presence of a gap at a corresponding time in the playback signal, and may be a combination of candidate noise estimates weighted by the gap confidence values. Generation of the candidate noise estimates may but need not include performance of echo cancellation. Optionally, noise compensation is performed on an audio input signal using the generated background noise estimate. Other aspects are systems configured to perform any embodiment of the noise estimation method.

An event sensing system includes one or more sensor assemblies, each sensor assembly including a housing, a microphone, and an audio signal processor. The audio signal processor is configured to generate one of an event classification, which is indicative of a type of event, and an event characteristic, which is indicative of a severity of an event, from the audio signal. The sensor assembly is configured to transmit event records that contain the generated event data to a cloud infrastructure component. The cloud infrastructure component is configured to calculate the probability of an event occurring at a plurality of locations in a region within a bounded time window based on the event records received from a plurality of sensor assemblies.