Creative Technology Corporation

Audience, Inc.

Knowles Electronics, Inc.

France Telecom

Creative Technologies Ltd.

Jean Laroche

Michael M Goodwin

Carlos M Avendano

Mark Barry Dolson

Dana C Massie

Jean-Marc Jot

Robert Sussman

Carlo Murgia

Martin Walsh

David P Rossum

Ludger Solbach

Sridhar Krishna Nemala

Mark Phillips

Michael Chorn

Wong Hoo Sim

Arvindh Krishnaswamy

David Klein

Brian D Clark

Ye Jiang

Leonardo Rub

Peter Santos

Alan W Peevers

Luke Dahl

Juha Merimaa

Umit Yapanel

Hong You

Jordan Cohen

Tyler Brown

A digital signal is processed by splitting it into at least two frequency subbands and the two subband signals are downsampled. A filter is applied in at least one of the subband signals. At least one of the phase and magnitude of the subband filtered signals is matched in the transition frequency band between the two subbands.

Alias-free subband processing

Described are noise suppression techniques applicable to various systems including automatic speech processing systems in digital audio pre-processing. The noise suppression techniques utilize a machine-learning framework trained on cues pertaining to reference clean and noisy speech signals, and a corresponding synthetic noisy speech signal combining the clean and noisy speech signals. The machine-learning technique is further used to process audio signals in real time by extracting and analyzing cues pertaining to noisy speech to dynamically generate an appropriate gain mask, which may eliminate the noise components from the input audio signal. The audio signal pre-processed in such a manner may be applied to an automatic speech processing engine for corresponding interpretation or processing. The machine-learning technique may enable extraction of cues associated with clean automatic speech processing features, which may be used by the automatic speech processing engine for various automatic speech processing.

Noise suppression for speech processing based on machine-learning mask estimation

Methods for voice sensing and keyword analysis are provided. An example method allows for causing a mobile device to transition to a second power mode, from a first power mode, in response to a first acoustic signal. The method includes authenticating a user based at least in part on a second acoustic signal. While authenticating the user, the second acoustic signal is compared to a spoken keyword. The spoken keyword is analyzed for authentication strength based on the length of the spoken keyword, quality of a series of phonemes used to represent the spoken keyword, and likelihood of the series of phonemes to be detected by a voice sensing. While receiving the first and second acoustic signals, a signal to noise ratio (SNR) is determined. The SNR is used to adjust sensitivity of a detection threshold of a voice sensing.

Voice sensing and keyword analysis

Noise suppression information is used to optimize or improve automatic speech recognition performed for a signal. Noise suppression can be performed on a noisy speech signal using a gain value. The gain to apply to the noisy speech signal is selected to optimize speech recognition analysis of the resulting signal. The gain may be selected based on one or more features for a current sub band and time frame, as well as one or more features for other sub bands and/or time frames. Noise suppression information can be provided to a speech recognition module to improve the robustness of the speech recognition analysis. Noise suppression information can also be used to encode and identify speech.

Noise suppression assisted automatic speech recognition

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.

Monaural noise suppression based on computational auditory scene analysis

The present technology provides techniques for transform domain reconstruction of noise-corrupted portions of an acoustic signal to emulate speech which is obscured by the noise. Replacement transform values for the noise-corrupted portions are determined utilizing the portions of the acoustic signal which contain speech.

Spectrum reconstruction for automatic speech recognition

Systems and methods described herein provide for low latency active noise cancellation which alleviate the problems associated with analog filter circuitry. The present technology utilizes low latency digital signal processing techniques which overcome the high latency conventionally associated with conversion between the analog and digital domains. As a result, low latency active noise cancellation is performed utilizing digital filter circuitry which is not subject to the inaccuracies and drift of analog filter components. In doing so, the present technology provides robust, high quality active noise cancellation.

Low latency active noise cancellation system

A system () for processing a sound signal () that allows dynamic customization of perceived spatial positions and sound qualities of sound components associated with the sound signal (). The system provides apparatus for processing a sound signal () that includes an input to receive the sound signal (), a sound unmixer () coupled to the input to receive the sound signal () and unmix at least one sound stream () from the sound signal () based on at least one unmixing instruction (), and an output coupled to the sound unmixer () to output the at least one sound stream ().

Process for removing voice from stereo recordings

A frequency band of an audio input signal is analyzed to determine if a transient is present. When transients are detected, modifications are made to the intensity levels corresponding to the frequency band for a brief time period.

Method for enhancing audio signals

Noise suppression is performed based on null coherence between sub-band signals of a primary acoustic signal and a secondary acoustic signal. The null coherence of a signal refers to portions of the signal that have high coherence and can be nullified by a null processor. The nullified component corresponds to target sources, such as an individual speaking into a phone. The coherence values indicate the presence of a target source and are used to suppress noise in portions of a signal that are not dominated by a desired target source. The inter-microphone level difference may be used in combination with the null coherence to provide noise suppression.

Noise suppression based on null coherence

Digital methods and systems for signal processing and filtering are provided. The methods and corresponding systems provide asynchronous conversion of sampling rate frequencies and utilize advanced multistage phasor filters for converting an input signal having a first sampling rate into an output signal sampled in an arbitrary sequence of sampling times. The conversion process provides a sequence of sets of complex numbers representing a filtered version of the input signal. More specifically, the conversion process includes the calculation of values of the output signal by multiplying (e.g., scaling) the sets of complex numbers by a corresponding set of complex phasors, the complex phasors corresponding to the timing of the arbitrary time sequence to obtain a corresponding set of real results with the value of the output signal being the sum of the real results.

Sample rate conversion using infinite impulse response filters

Robust feedforward active noise cancellation is provided which can overcome or substantially alleviate problems associated with the diverse and dynamic nature of the surrounding acoustic environment. A multi-faceted analysis is performed to determine the direction (or directions) of propagation of noise in the surrounding acoustic environment. The direction of propagation is then utilized to determine direction-dependent characteristics of the acoustic path between a reference position where the noise is captured and a desired position where the noise is to be cancelled. These characteristics are used to form a feedforward signal adapted to cancel the noise at the desired position. By forming the feedforward signal based on direction-dependent characteristics of the acoustic path, the techniques described herein can achieve optimal noise cancellation at the desired location, regardless of the direction of propagation of the noise.

Direction-aware active noise cancellation system

Systems and methods described herein provide for low latency active noise cancellation, which alleviates the problems associated with analog filter circuitry. The present technology utilizes low latency digital signal processing techniques that overcome the high latency conventionally associated with conversion between the analog and digital domains. As a result, low latency active noise cancellation is performed utilizing digital filter circuitry which is not subject to the inaccuracies and drift of analog filter components. In doing so, the present technology provides robust, high quality active noise cancellation.

Growing community of inventors

Santa Cruz, CA, United States of America

Jean Laroche