Provided is a signal processing apparatus including an acquisition unit that acquires an acoustic characteristic in a user's ear, isolated from the outside world and an NC filter unit that generates sound data having a phase opposite to an ambient sound leaking into the user's ear. The signal processing apparatus further includes a correction unit that corrects the sound data by using a correction filter and a determination unit that determines a filter coefficient of the correction filter based on the acoustic characteristic.

Systems and methods are provided an audio signal enhancement system that attenuates platform fan noise. Fan noise is a common type of self-noise in laptops and other devices, and fan noise can significantly degrade the quality of audio captured by built-in microphones. A neural network model is provided that enhances microphone Signal-to-Noise Ratio (SNR) and Signal-to-Distortion-plus-Noise Ratio (SDNR). The systems and methods also reduce algorithmic latency. The model architecture includes a Recurrent Neural Network, and a custom Gated Recurrent Unit layer is provided that uses fewer unique matrix weights and fewer biases and has fewer compute operations using fewer parameters. A platform self-noise suppression system is provided that eliminates low-amplitude platform self-noise signals. The model can predict when the platform fan is active, and remove the platform noise. In some examples, when the model predicts that the platform fan is not active, the model focuses on removing microphone self-noise.

The present invention provides a device for actively cancelling a target sound wavefront in an open space, the device comprising a signal processing module comprising at least one processor operatively coupled with a datastore, the at least one processor configured to: receive a data comprising one or more geographical features, and one or more audio features generated by one or more receiving microphones having a geographical relationship with an array of receiving microphones in an area adjacent to a user; process aid data using a prediction model adapting a trained deep learning framework; and provide output the inverse sound wavefront of the target sound at the area of said predicting microphones.

Examples of the disclosure describe systems and methods for reducing audio effects of fan noise, specifically, for a wearable system. A method wherein operating a fan of a wearable head device; detecting, with a microphone of the wearable head device, noise generated by the fan; generating a fan reference signal, wherein the fan reference signal represents at least one of a speed of the fan, a mode of the fan, a power output of the fan, and a phase of the fan; deriving a transfer function based on the fan reference signal and based further on the detected noise of the fan; generating a compensation signal based on the transfer function; and while operating the fan of the wearable head device, outputting, by a speaker of the wearable head device, an anti-noise signal, wherein the anti-noise signal is based on the compensation signal.

An active noise cancellation system can include an input node for receiving a noise signal, an output node for providing an anti-noise signal, and one or more blocks each configured to provide a transfer function between its input and output, with the block being further configured to be capable of having its transfer function estimated by a deep neural network framework. Such an active noise cancellation system can be supported by a computation engine that can be implemented on a system-on-chip device which in turn can be included in an audio device such as a headset.

Provided is a noise processing method, relating to the field of data processing. The method includes: obtaining a first noise signal at a position of a target person in a target workshop in a first time period; predicting an overall control parameter of multiple signal transmitters in a second time period based on the first noise signal to obtain a current parameter prediction result; determining one or more target transmitters that need to work in the second time period from the multiple signal transmitters based on the result, and obtaining a parameter prediction value of each target transmitter in the second time period; and controlling each target transmitter to transmit a noise interference signal in the second time period according to a corresponding parameter prediction value to weaken a second noise signal at the position of the target person in the target workshop in the second time period.

Automated attention handling techniques are described herein for use with wearable audio components with active noise control (ANC) to suppress ambient sound. A name embedding model is trained automatically to convert name audio samples into acoustic segments based on a knowledge distillation model. The name embedding model is used to generate reference embeddings for each of a user-enrolled set of names, and a relation network and a false rejection network are also trained. In real-time operation, the name embedding model converts real-time audio samples to real-time embeddings, the relation network compared the real-time embeddings to the reference embeddings to look for candidate matches, and the false rejection network validates the candidate matches to detect when one of the user-enrolled names has been invoked. Detecting such an invocation automatically triggers the ANC to switch to a conversation mode.

Techniques for automatic parameter tuning of active road noise cancellation systems are described herein. The system can automatically search for an optimal set of algorithm parameters based on recorded data. An active road noise cancellation algorithm and simulation can be embedded in an auto-differentiation framework, which allows gradients of the algorithm parameters to guide the automatic search and calculations of the algorithm parameters.

A vehicle noise cancelling apparatus including: at least one microphone provided on a vehicle and configured to detect a noise generated from one or more noise sources; at least one antiphase speaker configured to generate an antiphase sound wave; one or more processors; and one or more memories storing program instructions, where, by executing the program instructions, the one or more processors are configured to: obtain the detected noise from the at least one microphone, measure parameters based on an environment in which the vehicle is being driven, determine a radiation pattern of the noise based on the detected noise and the measured parameters, produce a sound wave in antiphase to the detected noise based on the radiation pattern of the noise, and control the at least one antiphase speaker to generate the antiphase sound wave.

System and methods are provided for estimating the sound field from partial observations. Estimating an acoustic environment for virtual reality and augmented reality applications is a step in the creation of simulated acoustic sound scenes. In particular, the impulse responses of room can be estimated with a generative model. In a teleconferencing scenario with remote participants and a group of participants in a common physical space, giving the remote participants the impression that all other participants are sitting is in the same room acoustically requires filtering the speech of the remote participants with impulse responses estimated at the desired rendering position in the conference room.