Abstract: An artificial intelligence (AI) device may acquire a probability that a received speech signal is classified as a noise signal, calculate a confidence level of a first model for determining to which phoneme the speech signal belongs, based on the speech signal, determine a weight of the first model based on the probability and the confidence level of the first model, and output a speech recognition result of the speech signal using the determined weight of the first model.

Abstract: Methods and devices for processing and voice operated control are provided. The method can include performing a non-difference comparison between a first received sound and a second received sound, determining if speech exists based on the comparison, and transmitting or providing a decision that the speech is present to at least one among the device, a cell phone, a media player, or a portable computing device. Other embodiments are disclosed.

Abstract: Methods and systems for advanced stereo processing of an audio signal are disclosed. The methods and systems include selecting a coding mode of either transform coding or linear predictive coding and performing advanced stereo processing when in the selected coding mode. Both encoding and decoding operations are provided.

Abstract: An active noise control device includes: a control target signal extractor for extracting a signal component of a control target frequency from an error signal as a control target signal which is a complex-valued signal having a real part and an imaginary part; a control signal generator for generating a control signal for controlling a control actuator, by signal-processing the control target signal through a control filter; and a control filter coefficient updater for successively and adaptively updating the coefficient of the control filter.

Abstract: Systems and methods are provided for measuring the distortion and muffling caused by a face mask. For example, in one embodiment a simulated voice source produces a sound. The sound is then acoustically coupled to a simulated vocal tract and a face mask. A microphone receives sound and produces a signal and an analyzer receives the signal from the microphone. A manikin head or other facial structure may also simulate fitting of the face mask onto a face. The analyzer may further produce a quantitative assessment of the distortion and muffling of the face mask, for example, by comparing at least one spectrum obtained with the face mask and at least one spectrum obtained without the face mask.

Abstract: Embodiments of the present disclosure relate to object detection in an image. In an embodiment, a computer-implemented method is disclosed. According to the method, image data representing a scene is obtained and sound distribution information related to the scene is obtained. A detection strategy to be applied in object detection is determined based on the sound distribution information. The object detection is performed on the image data by applying the detection strategy. In other embodiments, a system and a computer program product are disclosed.

Abstract: A method, apparatus, and computer-readable storage medium that modulate a composition of an audio output in accordance with a noise level of an environment. For instance, the present disclosure describes a method for modulating an audio output of a microphone array, comprising receiving two or more audio signals from two or more microphone capsules in the microphone array, each audio signal comprising an electrical noise of a corresponding microphone capsule and a response to acoustic stimuli in an environment perceived by the microphone capsule, estimating an acoustic contribution level of the environment based on the received audio signals, and determining, by processing circuitry, a composition of the audio output of the microphone array based on the estimated acoustic contribution level of the environment, the composition being based on at least a relationship between acoustic noise and directivity indices of each of a plurality of beamformers.

Abstract: A method includes selecting a frame of an audio signal. The method further includes determining a first power spectral density (PSD) distribution of the frame. The method further includes generating a first reference PSD distribution indicating an estimate of background noise in the frame based on a non-linear weight, a second reference PSD distribution of a previous frame of the audio signal, and a second PSD distribution of the previous frame. The method further includes determining whether voice activity is detected in the frame based on the first PSD distribution of the frame and the first reference PSD distribution.

Abstract: Methods, systems, and devices for signal processing are described. Generally, as provided for by the described techniques, a wearable device to receive an input audio signal from one or more outer microphones, an input audio signal from one or more inner microphones, and a bone conduction signal from a bone conduction sensor based on the input audio signals. The wearable device may filter the bone conduction signal based on a set of frequencies of the input audio signals, such as a low frequency portion of the input audio signals. For example, the wearable device may apply a filter to the bone conduction signal that accounts for an error in the input audio signals. The wearable device may add a gain to the filtered bone conduction signal and may equalize the filtered bone conduction signal based on the gain. The wearable device may output an audio signal to a speaker.

Abstract: Headphone providing fully natural interface are described. According to one aspect of such headphones, the headphones comprises a microphone configured for capturing an ambient sound, a speaker configured for playing audio signals, a command interface configured for receiving one or more external control commands, and a control unit having an ambient sound monitoring function. The control unit captures the ambient sound through the microphone, and automatically causes the headphones to enter an interactive mode when a preset interested sound is detected to appear in the ambient sound. The control unit controls the headphones to output an interactive reminder in the interactive mode, and the interactive reminder comprises one or more of visual reminders, a tactile reminder and an auditory reminder. Thus, interaction between the user and the ambience can be realized in a fully natural interface manner according to user preferences.

Abstract: Systems, methods, and devices are disclosed for detecting an active speaker in a two-way conference. Real time audio in one or more sub band domains are analyzed according to an echo canceller model. Based on the analyzed real time audio, one or more audio metrics are determined from output from an acoustic echo cancellation linear filter. The one or more audio metrics are weighted based on a priority, and a speaker status is determined based on the weighted one or more audio metrics being analyzed according to an active speaker detection model. For an active speaker status, one or more residual echo or noise is removed from the real time audio based on the one or more audio metrics.

Abstract: Invoking one or more previously dormant functions of an automated assistant in response to detecting, based on processing of vision data from one or more vision components: (1) a particular gesture (e.g., of one or more “invocation gestures”) of a user; and/or (2) detecting that a gaze of the user is directed at an assistant device that provides an automated assistant interface (graphical and/or audible) of the automated assistant. For example, the previously dormant function(s) can be invoked in response to detecting the particular gesture, detecting that the gaze of the user is directed at an assistant device for at least a threshold amount of time, and optionally that the particular gesture and the directed gaze of the user co-occur or occur within a threshold temporal proximity of one another.

Abstract: The present disclosure provides new variants of non-negative matrix factorization suitable for separating desired audio content from undesired audio content. In certain embodiments, a multi-dimensional non-negative representation of an audio signal is decomposed into desired content and undesired content by performing convolutional non-negative matrix factorization (CNMF) on multiple layers, each layer having a respective non-negative matrix representation. In certain embodiments, the desired content is represented by a first dictionary and the undesired content is represented by a second dictionary, and sparsity is imposed on activations of basic elements of the first or the second dictionary, wherein a degree of sparsity is controlled by setting a minimum number of components with significant activations of the first or second dictionary, respectively.

Abstract: The present disclosure relates to a method, apparatus, and computer-readable storage medium that modulate an audio output of a microphone array in order to isolate speech of a talker in a vehicle. For instance, the present disclosure describes a method for modulating an audio output of a microphone array, comprising receiving two or more audio signals from two or more microphone capsules in the microphone array, each audio signal comprising a response to acoustic stimuli in an environment perceived by a respective microphone capsule of the microphone array, estimating an acoustic noise contribution level of the environment based on the received audio signals, estimating a voice contribution level of the environment based on the received audio signals, and determining, by processing circuitry, a direct voice contribution level of the environment based on the estimated acoustic noise contribution level of the environment and the estimated voice contribution level of the environment.

Abstract: A method operates a hearing device. The hearing device has a microphone by which ambient sound is picked up and is converted into an input signal that has a wanted component and a noise component. A stationarity of the input signal is determined. A signal-to-noise ratio of the input signal is determined on a basis of a scaling factor. The scaling factor is determined on a basis of the stationarity, namely on a basis of a function that indicates the scaling factor on a basis of the stationarity of the input signal. A corresponding hearing device implements such a method.

Abstract: Tampering with an audio/video (A/V) recording and communication device is detected based on audio data captured by a microphone and/or video data captured by a camera of the A/V recording and communication device. The detection of the tampering may be based on, for example, processing of the audio and/or video data. Additional data may be collected and/or other actions taken in response to detection of the tampering.

Abstract: An information providing device to be mounted on a vehicle. The vehicle includes control means and control means for transmitting, when the vehicle is controlled by automatic driving, a control signal for outputting route guidance information for informing a driver of the vehicle about a specified route by mechanical speech, and a control signal for outputting automatic driving guidance information for informing the driver of the vehicle about a route of the vehicle driven by automatic driving by a sound effect.

Abstract: Aspects of the present disclosure relate to distinguishing voice commands. One or more stored blocked directions of background voice noise from one or more audio output devices for a location of a voice command device are accessed. A voice input is received at the voice command device at the location and a determination is made that the voice input is received from a blocked direction. A status of an audio output device is queried to determine whether it is emitting audio. In response to a determination that the audio output device is currently emitting audio, an audio file is obtained from the audio output device, the audio file corresponding to a time when the voice input was received. The obtained audio file is compared with the received voice input. The received voice input is ignored if there is a substantial match with the obtained audio file.

Abstract: A method for processing audio data includes obtaining a first noise frame of an audio signal, wherein the first noise frame includes a first low-band signal and a first high-band signal, obtaining a first low-band parameter corresponding to the first low-band signal and a first high-band parameter corresponding to the first high-band signal, encoding a first silence insertion descriptor (SID) corresponding to the first noise frame to comprise the first low-band parameter and the first high-band parameter, obtaining a second noise frame of the audio signal, wherein the second noise frame includes a second low-band signal and a second high-band signal, where the first noise frame is prior to the second noise frame in the audio signal, and determining whether a second SID corresponding to the second noise frame should comprise a second high-band parameter of the second high-band signal.

Abstract: A voice-interaction system enables mobile workers to capture measurements, observations and complete inspections using their voice as they move about, leaving one or both hands and eyes free to safely and effectively focus on work tasks. The system has the flexibility to recognize highly specialized vocabulary, prompting for and error-checking utterances that are unique to an industry, company, government agency, user or specific task. This data is saved and formatted to be viewed, listened to, or input into a structured data-base for further use.

Abstract: A system has multiple audio-enabled devices that communicate with one another over an open microphone mode of communication. When a user says a trigger word, the nearest device validates the trigger word and opens a communication channel with another device. As the user talks, the device receives the speech and generates an audio signal representation that includes the user speech and may additionally include other background or interfering sound from the environment. The device transmits the audio signal to the other device as part of a conversation, while continually analyzing the audio signal to detect when the user stops talking. This analysis may include watching for a lack of speech in the audio signal for a period of time, or an abrupt change in context of the speech (indicating the speech is from another source), or canceling noise or other interfering sound to isolate whether the user is still speaking.

Abstract: Embodiments of the present disclosure set forth a method of decomposing an audio signal into a set of sub-band signals and detecting a set of signal energy values, where each signal energy value is associated with a sub-band signal. The method also includes generating a noise reduction threshold based on at least one sub-band signal, and, for each sub-band signal, comparing the associated signal energy value to the noise reduction threshold. Based on determining that at least one sub-band signal is associated with a signal energy value below the noise reduction threshold, the method includes attenuating the at least one the sub-band signal to generate a set of attenuated sub-band signals. The method also includes combining at least one sub-band signal included in the set of sub-band signals with at least one attenuated sub-band signal included in the set of attenuated sub-band signals to generate an output audio signal.

Abstract: A hearing device includes: an input module for provision of a first input signal; a processor configured to provide an electrical output signal based on the first input signal; a receiver configured to provide an audio output signal; and a controller comprising a speech intelligibility estimator configured to determine a speech intelligibility indicator indicative of speech intelligibility based on the first input signal, wherein the controller is configured to control the processor based on the speech intelligibility indicator; wherein the speech intelligibility estimator comprises a decomposition module configured to decompose the first input signal into a first representation of the first input signal in a frequency domain, wherein the first representation comprises one or more elements representative of the first input signal; and wherein the decomposition module comprises one or more characterization blocks for characterizing the one or more elements of the first representation in the frequency domain.

Abstract: The invention addresses the problem of recovering an unknown signal from multiple records of brief duration which are presumed to contain the signal at mutually random delays in a background of independent noise. The scenario is relevant to many applications, among which are the recovery of weak transients from large arrays of sensors and the identification of recurring patterns through a comparison of sequential intervals within a single record of longer duration. A simple and practical approach is provided by solving this problem through higher-order spectra. Applying the method to the third-order spectrum, the bispectrum, leads to filters derived from cross bicoherence.

Abstract: A system and method are described for smart noise cancellation. In some implementations, the system may receive audio data describing an audio signal, which the system may use to determine a set of frames of the audio signal. Spectral analysis, which may include a signal-to-noise ratio estimate, may be performed on the one or more frames of the audio. In some instances, the system may identify a noise frame from among the one or more frames based on the spectral analysis, and may reduce noise in the one or more frames based on the noise frame and the spectral analysis on the one or more frames of the audio signal.

Abstract: A method for filtering a sound signal acquired by a voice recognition system is proposed wherein the filter used to delete the unwanted sound signal is based on a noise model created from a history of preceding acquisitions of the sound signal. The method is based on the history for predicting the characteristics of the signal in the course of acquisition and relies on this prediction to establish the noise model used by the filtering function.

Abstract: A log spectral envelope sequence L0, L1, . . . , LN?1 and an envelope code for the log spectral envelope sequence L0, L1, . . . , LN?1 are obtained. The log spectral envelope sequence L0, L1, . . . , LN?1 is an integer value sequence corresponding to binary logarithms of respective sample values of a spectral envelope sequence and is an integer value sequence whose total sum is 0. For a quantized spectral sequence {circumflex over (?)}X0, {circumflex over (?)}X1, . . . , {circumflex over (?)}XN?1, a smoothed spectral sequence ˜X0, ˜X1, . . . , ˜XN?1 is obtained by: for {circumflex over (?)}Xk with Lk being a positive value, adopting {circumflex over (?)}Xk with Lk digits from its least significant digit removed as ˜Xk; for {circumflex over (?)}Xk with Lk being a negative value, adopting {circumflex over (?)}Xk with ?Lk digits added to its least significant digit in accordance with a predefined rule as ˜Xk; and when Lk is 0, adopting {circumflex over (?)}Xk as ˜Xk.

Abstract: A method for encoded-sound determination performed by a computer includes: executing a first process that includes obtaining information indicating intensities of sound signals, the frequencies being calculated from the sound signals and corresponding to frequencies; and executing a second process that includes determining whether or not the sound signals are signals of encoded sound, based on whether or not the intensities of the sound signals in predetermined frequency bands that are adjacent to each other in a frequency direction have a difference that is larger than or equal to a predetermined threshold.

Abstract: A mechanism is described for facilitating wind detection and wind noise reduction in computing environments according to one embodiment. An apparatus of embodiments, as described herein, includes wind detection logic to detect wind associated with the apparatus including a wearable computing device, wherein the wind is detected based on samples from multiple microphones and extraction and use of multiple features including spectral sub-band centroid (SSC) features and coherence features; and decision and execution logic to reduce wind noise associated with the detected wind.

Abstract: A method includes generating, by a noise generator of an apparatus, a noise signal stream. The method includes outputting, via an audio output device of the apparatus, the noise signal stream towards an audio input device of a personal communication device. The noise signal stream travels from the audio output device towards the audio input device via an audio-sealing pathway. The method includes receiving a recorded representation of the noise signal stream from the personal communication device. The recorded representation of the noise signal stream is a function of a transformation of the noise signal stream, by the audio-sealing pathway, and characterizes an audio transfer function of the audio-sealing pathway. The method includes determining, by the controller of the apparatus, whether or not the audio-sealing pathway satisfies an operational criterion based on the recorded representation of the noise signal stream and a reference transfer function associated with the audio-sealing pathway.

Abstract: The present disclosure provides a speech enhancement method and apparatus, a device and a storage medium. The method includes: acquiring a first speech signal and a second speech signal; obtaining a signal to noise ratio of the first speech signal; determining, according to the signal to noise ratio of the first speech signal, a fusion coefficient of filtered signals corresponding to the first speech signal and the second speech signal; and performing, according to the fusion coefficient, speech fusion processing on the filtered signals corresponding to the first speech signal and the second speech signal to obtain an enhanced speech signal. Thereby, it is realized that a fusion coefficient of speech signals of a non-air conduction speech sensor and an air conduction speech sensor is adaptively adjusted according to environment noise, thereby improving the signal quality after speech fusion, and improving the effect of speech enhancement.

Abstract: An apparatus for processing an audio signal includes an audio signal analyzer and a filter. The audio signal analyzer is configured to analyze an audio signal to determine a plurality of noise suppression filter values for a plurality of bands of the audio signal, wherein the analyzer is configured to determine a noise suppression filter value so that a noise suppression filter value is greater than or equal to a minimum noise suppression filter value and so that the minimum noise suppression value depends on a characteristic of the audio signal. The filter is configured for filtering the audio signal, wherein the filter is adjusted based on the noise suppression filter values.

Abstract: Method and apparatus are disclosed for speech recognition for vehicle voice commands. An example vehicle includes a microphone to collect a signal including a voice command, memory, and a controller. The controller is configured to determine an initial identification by feeding the signal into a first automatic speech recognition (ASR) engine and determine habits by feeding user history into a habits engine. The controller also is configured to identify the voice command by feeding the signal, the initial identification, and the habits into a second ASR engine. The controller also is configured to perform a vehicle function based on the voice command.

Abstract: A speech/audio bitstream decoding method includes acquiring a speech/audio decoding parameter of a current speech/audio frame, where the foregoing current speech/audio frame is a redundant decoded frame or a speech/audio frame previous to the foregoing current speech/audio frame is a redundant decoded frame, performing post processing on the acquired speech/audio decoding parameter according to speech/audio parameters of X speech/audio frames, where the foregoing X speech/audio frames include M speech/audio frames previous to the foregoing current speech/audio frame and/or N speech/audio frames next to the foregoing current speech/audio frame, and recovering a speech/audio signal using the post-processed speech/audio decoding parameter of the foregoing current speech/audio frame. The technical solutions of the speech/audio bitstream decoding method help improve quality of an output speech/audio signal.

Abstract: A signal processing apparatus includes a detection unit configured to perform a voice detection process on each of a plurality of audio signals captured by a plurality of microphones arranged at mutually different positions, a determination unit configured to determine a degree of similarity between two or more of the plurality of audio signals in which voice is detected by the detection unit, and a suppression unit configured to perform a process of suppressing the voice contained in at least one of the two or more audio signals, in response to a determination that the degree of similarity between the two or more audio signals is less than a threshold by the determination unit.

Abstract: Disclosed herein is a method for RNN-based noise reduction in a real-time conference, comprising: performing frame-and-window for a speech signal to obtain a logarithmic spectrum of the speech signal, and placing the logarithmic spectrum into the RNN model to determine a noise reduction suppression coefficient, and then obtaining the denoised speech signal by applying the noise reduction suppression coefficient to the logarithmic spectrum of the original signal, thereby achieving utilization of the RNN noise reduction method in real-time conferences. In the present disclosure, when inputting the RNN model for estimation, only the logarithmic spectrum of the current frame needs to be inputted. The RNN model of the present disclosure has few requirements on inputted information, without performing huge preprocessing on the received speech signal, which in turn reduces computation burden, increases response speed, and enhances real-time performance.

Abstract: An encoder for providing an audio stream on the basis of a transform-domain representation of an input audio signal includes a quantization error calculator configured to determine a multi-band quantization error over a plurality of frequency bands of the input audio signal for which separate band gain information is available. The encoder also includes an audio stream provider for providing the audio stream such that the audio stream includes information describing an audio content of the frequency bands and information describing the multi-band quantization error. A decoder for providing a decoded representation of an audio signal on the basis of an encoded audio stream representing spectral components of frequency bands of the audio signal includes a noise filler for introducing noise into spectral components of a plurality of frequency bands to which separate frequency band gain information is associated on the basis of a common multi-band noise intensity value.

Abstract: Systems and methods are disclosed herein for modifying modulated signals for transmission. The system receives a modulated signal comprising a speech signal and a carrier wave and generates first and second spectral signals by converting the modulation signal and carrier wave from the time domain to the frequency domain respectively. The system then determines spectral bands for the first and second spectral signals. For each spectral band, the system calculates a weighted spectral band value based on a magnitude of the first spectral signal within the spectral band and generates a modified spectral signal by modifying the second spectral signal with the weighted spectral band value. The system then converts the modified spectral signal from the frequency domain to the time domain and transmits the converted modified spectral signal to a server.

Abstract: The present disclosure discloses a method for processing a voice in interior environment of a vehicle, an electronic device and a storage medium. The method includes the following. A reference audio is acquired, and the reference audio is recorded to obtain a recorded audio. A pure voice is acquired. Noise data for each part or period of the recorded audio satisfying a target signal-to-noise ratio condition pertaining to that part is selected from the recorded audio, and the noise data is superimposed to the pure data to obtain a noisy voice. The noisy voice and the reference audio are inputted to an acoustic echo canceller (AEC) module as inputted data. The AEC module is configured to perform an echo cancellation operation on the inputted data to obtain training data having AEC residual noise.

Abstract: An active sound barrier has at least one passive sound absorber at or near a boundary location. A microphone provides an output to a frequency division module, in which a plural of frequencies are filtered to provide outputs corresponding to frequency segments of the receiving transducer output at respective ones of the frequencies. An active driving circuit drives plural speakers or output transducers at respective ones of the frequencies, with at least a subset of the speakers or output transducers at or close to the barrier. The speakers or output transducers cooperate with the passive sound absorber to reduce noise across a wide frequency band as well as to effect an electrically switchable soft boundary.

Abstract: Disclosed are a headset and an operating method thereof, which execute a mounted artificial intelligence (AI) algorithm and/or machine learning algorithm and communicate with different electronic apparatuses and external servers in a 5G communication environment. The headset includes a plurality of microphones, an AEC, a VAD, and a BSS, which process sound signals received from the microphones, resulting in improved convenience for a user.

Abstract: A computer-implemented method according to one embodiment includes creating a clean dictionary, utilizing a clean signal, creating a noisy dictionary, utilizing a first noisy signal, determining a time varying projection, utilizing the clean dictionary and the noisy dictionary, and denoising a second noisy signal, utilizing the time varying projection.

Abstract: A headset adapted to transmitting an outgoing audio signal (Tx) and comprising a voice microphone and at least one ambient microphone. The headset comprises an ambient noise reduction block (ANR), which is adapted to reduce the level of ambient noise in the outgoing audio signal (Tx). The ambient noise reduction block (ANR) includes the measuring the levels of the voice microphone signal (X) and ambient microphone signal (Y) to estimate a characteristic constant level drop between the voice microphone signal (X) and the ambient microphone signal (Y), which is characteristic for the headset user talking. A time-varying filter passes the voice microphone signal (X), when the level difference is larger than characteristic constant level drop (CLD) and attenuates the voice microphone signal (X), if the level difference is below characteristic constant level drop.

Abstract: A method begins by a processing module dispersed storage error encoding a data segment to produce a set of encoded data slices and dispersed storage error encoding metadata associated with the data segment to produce a set of encoded metadata slices. The method continues with the processing module creating a set of data slice names for the set of encoded data slices and creating a set of metadata slice names based on the set of data slice names. The method continues with the processing module sending the set of encoded data slices and the set of data slice names to a dispersed storage network (DSN) memory for storage therein. The method continues with the processing module sending the set of encoded metadata slices and the set of metadata slice names to the DSN memory for storage therein.

Abstract: The invention provides an earphone and a set of earphones. The earphone includes a processing circuit and a filtering module. The processing circuit acquires a first speech signal and performs a pre-processing operation on the first speech signal to generate a second speech signal. The filtering module includes high-pass, low-pass, and band-pass filters. The processing circuit is further configured to: receive first, second, and third signals respectively from the high-pass, low-pass, and band-pass filters; perform a noise reduction operation on the second and third signals to generate a fourth signal; and perform a signal synthesis operation on the first and fourth signals to synthesize the first and fourth signals to form an output speech signal.

Abstract: An adaptive beamformer includes at least first and second microphones that generate respective audio signals that include speech and noise, a controller that detects occurrences of speech and noise within the audio signals, an adaptive speech cancelling filter that cancels speech from the audio signal of the second microphone to provide a speech-cancelled signal, an adaptive mixing block that combines the speech-cancelled signal and the second microphone audio signal to provide a noise reference signal in a weighted manner such that a weight of the second microphone signal is increased proportionally with an amount of the detected noise and a weight of the speech-cancelled signal is increased proportionally with an amount of the detected speech, and an adaptive noise cancelling filter that uses the noise reference signal to remove the noise from the first microphone audio signal.

Abstract: Provided are, among other things, systems, methods and techniques for processing an audio signal to add virtual bass. In one representative embodiment, an apparatus includes: an input line that inputs an original audio signal; an estimator, coupled to the input line, that estimates a fundamental frequency of a bass sound within the original audio signal; a bass extraction filter, coupled to the input line, that extracts a bass portion of the original audio signal that is at least 1 octave wide and includes the fundamental frequency; a frequency translator, coupled to the bass extraction filter, that shifts the bass portion, in its entirety, by a positive frequency increment that is an integer multiple of the fundamental frequency estimated by the estimator, thereby providing a virtual bass signal; and an adder having 1) inputs coupled to the original audio signal and to the virtual bass signal and 2) an output.

Abstract: The present disclosure discloses a method and an apparatus for judging termination of sound reception and a terminal device. The method including: performing a voice activity detection on a current sound clip to obtain a first value; performing a semantic relevance detection on the current sound clip and a next sound clip by deep learning to obtain a second value; performing a weighted calculation on the first value and the second value to obtain a third value; comparing the third value with a preset threshold; and determining whether sound reception of the current sound clip is terminated based on the comparison result.

Abstract: A process for compressing an audio speech signal utilizes ASR processing to generate a corresponding text representation and, depending on confidence in the corresponding text representation, selectively applies more, less, or no compression to the audio signal. The result is a compressed audio signal, with corresponding text, that is compact and well suited for searching, analytics, or additional ASR processing.

Abstract: Systems, devices, and methods are described for reducing degradation of a voice recognition input. An always listening device may always be listening for voice commands via a microphone and may experience interference from unwanted audio such as from the output audio of television speakers. The always listening device may receive data associated with the output audio over a first communications channel. The always listening device may also receive, on a second communications channel, timing information associated with data. The always listening device may adjust admission of the audio received by the microphone to enable it to arrive at approximately the same time as the data received via the first communications channel. The unwanted output audio included in the audio received via the microphone may then be determined and may be removed so that a voice command in the audio received by the microphone may be processed.