Abstract: An audio encoder has a window function controller, a windower, a time warper with a final quality check functionality, a time/frequency converter, a TNS stage or a quantizer encoder, the window function controller, the time warper, the TNS stage or an additional noise filling analyzer are controlled by signal analysis results obtained by a time warp analyzer or a signal classifier. Furthermore, a decoder applies a noise filling operation using a manipulated noise filling estimate depending on a harmonic or speech characteristic of the audio signal.

Abstract: An audio signal may be received, in a processor associated with a vehicle. Sound related vehicle information representing one or more sounds may be received by the processor. The sound related vehicle information may or may not include an audio signal. A speech recognition process or system may be modified based on the sound related vehicle information.

Abstract: In an adaptive phase discovery system a first audio signal is received via a first microphone and a second signal is received via a second microphone. Corresponding audio frames of the first and second signals are each transformed into the frequency domain and a plurality of frequency sub-bands are generated. A phase is determined for each frequency sub-band in each signal. Instantaneous phase differences are determined between the signals at each of the frequency sub-bands. Lower frequency instantaneous phase differences are filtered over time to determine current phase differences at lower frequencies. When SNR is high in lower frequency sub-bands, lower frequency sub-band phase differences are tracked to the higher frequency sub-bands. The tracked higher frequency phase differences are filtered over time to determine phase differences for the current frame. The phase differences may be used to rotate phases in each sub-band and sum signals and/or to reject off-axis signals.

Abstract: A dynamic range compressor of a subband type for carrying out a dynamic compression on a broadband input signal includes a subband splitting device for splitting the broadband input signal into K narrowband subband signals. An amplifier unit amplifies each of the K subband signals to obtain K amplified subband signals. Further, a subband combining device is provided for combining the K amplified subband signals to obtain a broadband output signal, which is a dynamically compressed version of the broadband input signal. An envelope detecting device generates, for each of the K subbands, a respective one of K envelope signals. An amplifier control device generates, in dependence of the K envelope signals, K amplifier control signals, each being representative of one of the K amplification factors. The amplifier control device is adapted to generate an amplification control signal in dependence of more than one of the K envelope signals.

Abstract: Techniques are provided for vector noise cancellation. Different value combinations for a plurality of weighting factors may be established for a plurality of selection regions. Each value combination for the plurality of weighting factors may correspond to a different combination of a set of input signals. One or more characteristics of input signals may be used to select a particular selection region. A particular value combination of the set of weighting factors may be chosen to attenuate or amplify the input signals to generate one or more output signals.

Abstract: A disclosed adjustment apparatus includes: a calculation unit that calculates a ratio between a first frequency characteristic in a first frequency bandwidth of voice signals and a second frequency characteristic in a second frequency bandwidth of the voice signals, which is higher than the first frequency bandwidth, and calculates an adjustment amount for adjusting at least a portion of a frequency characteristic of the voice signals so that the calculated ratio approaches a predetermined reference, when the calculated ratio does not satisfy the predetermined reference; and a modification unit that modifies at least the portion of the frequency characteristic of the voice signals according to the adjustment amount.

Abstract: A direct sound extraction device includes: a spectrum transform unit that transforms an input signal, which includes a reverberant sound in a direct sound and on which a Fourier transform process has been performed, to a first amplitude spectrum signal Lfa; a low-pass filter unit (4) that performs a low-pass filtering process on the first amplitude spectrum signal Lfa for each frequency to generate a second amplitude spectrum signal Lfa1; a first subtraction unit (18) that calculates a third amplitude spectrum signal by subtracting the second amplitude spectrum signal Lfa1 from the first amplitude spectrum signal Lfa; and an inverse Fourier transform unit that generates a direct sound signal Lfd from a frequency spectrum signal calculated based on a phase spectrum signal and the third amplitude spectrum signal.

Abstract: Implementations of systems, method and devices described herein enable enhancing the intelligibility of a target voice signal included in a noisy audible signal received by a hearing aid device or the like. In particular, in some implementations, systems, methods and devices are operable to generate a machine readable formant based codebook. In some implementations, the method includes determining whether or not a candidate codebook tuple includes a sufficient amount of new information to warrant either adding the candidate codebook tuple to the codebook or using at least a portion of the candidate codebook tuple to update an existing codebook tuple. Additionally and/or alternatively, in some implementations systems, methods and devices are operable to reconstruct a target voice signal by detecting formants in an audible signal, using the detected formants to select codebook tuples, and using the formant information in the selected codebook tuples to reconstruct the target voice signal.

Abstract: A hearing aid comprising a time domain codec. The codec comprises a decoder adapted to generate a decoded output signal based on an input quantization index and an encoder for generating an output quantization index based on an input signal, said encoder comprising said decoder and a predictor receiving an excitation signal derived from said decoder output signal and outputting a prediction signal. The output quantization index is determined by repeated decoding of the quantization indices in order to minimize the error between the input signal and the prediction signal, and the predictor uses a recursive autocorrelation estimate for the error minimization. The invention further provides a method of encoding an audio signal.

Abstract: A method of controlling sounds associated with a vehicle is provided. The method includes performing on a processor, monitoring powertrain data; determining a powertrain transition event based on the powertrain data; and selectively controlling the generation of one or more tones based on the powertrain transition event.

Abstract: An apparatus comprising at least one processor and at least one memory including computer program code the at least one memory and the computer program code configured to, with the at least one processor, cause the apparatus at least to perform determining a signal identification value for an audio signal, determining at least one noise level value for the audio signal, comparing the signal identification value against a signal identification threshold and each of the at least one noise level value against an associated noise level threshold, and identifying the audio signal dependent on the comparison.

Abstract: A method of controlling sounds associated with a vehicle is provided. The method includes: performing on a processor, monitoring engine torque; and selectively controlling the generation of one or more tones associated with the vehicle based on the engine torque.

Abstract: A system and method for mitigating microphone hiss may obtain a frequency spectrum characteristic for a microphone. A microphone that has limited dynamic range may create microphone hiss in an output signal. The microphone hiss may prevent a reproduction of a sound field, represented in an output signal of the microphone, from being perceived as a natural environment. The microphone frequency spectrum may be obtained using static measurements or calculated dynamically. A virtual noise floor may be calculated responsive to the microphone frequency spectrum and a desired noise floor. Gain coefficients may be calculated responsive to the output signal of the microphone. The gain coefficients may be calculated to mitigate undesirable signal content including background noise and echoes. The calculated gain coefficients may be modified responsive to the virtual noise floor. The modified gain coefficients may allow a reproduction of the sound field to be perceived as a natural environment.

Abstract: A system for detecting noise in a signal received by a microphone array and a method for detecting noise in a signal received by a microphone array is disclosed. The system also provides for the reduction of noise in a signal received by a microphone array and a method for reducing noise in a signal received by a microphone array. The signal to noise ratio in handsfree systems may be improved, particularly in handsfree systems present in a vehicular environment.

Abstract: Embodiments of the disclosure relate to selective voice transmission and include receiving an identification of one or more authorized speakers for a telephone call and retrieving a voice sample for each of the one or more authorized speakers. Embodiments also include receiving one or more audio signals for the telephone call and filtering the one or more audio signals by removing a portion of the one or more audio signals that do not contain a voice of at least one of the one or more authorized speakers in the one or more audio signals.

Abstract: A clear picture of who is speaking in a setting where there are multiple input sources (e.g., a conference room with multiple microphones) can be obtained by comparing input channels against each other. The data from each channel can not only be compared, but can also be organized into portions which logically correspond to statements by a user. These statements, along with information regarding who is speaking, can be presented in a user friendly format via an interactive timeline which can be updated in real time as new audio input data is received.

Abstract: Among other things, a sound processing device system is disclosed to assist a hearing-impaired human listener recognize speech sounds or phonemes. The device system may be configured at least to generate an output audio signal at least by transposing and causing a negative rank ordering of frequency of at least a portion of the input audio signal. Compression also may be performed on the at least the portion of the input audio signal as part of generating the output audio signal. The negative rank ordering may be performed on a high-frequency portion of the input audio signal that becomes a low-frequency portion of the output audio signal by the transposing. The low-frequency portion of the output audio signal may represent an inverted ordering of frequencies or frequency segments present in the high-frequency portion of the input audio signal.

Abstract: There is provided a method of determining a transmission quality when receiving audio signals which are transmitted over a frequency-modulated path in analog form. For that purpose a frequency-modulated audio signal is received and IQ demodulation of the input signal is implemented. The I- and the Q-path of the demodulated baseband signal are subjected to analog/digital conversion. Alternatively the input signal can first be digitized and then subjected to IQ demodulation. A multiplicity of N samples is detected. The amount of the N samples is formed. A reference value is determined from the amount of the N samples. Each of the N samples is standardized to the reference value. A measurement in respect of the fluctuations in the standardized amounts is determined. The measurement in respect of the fluctuations indicates the quality of the transmission path.

Abstract: An audio-based system may perform automatic noise reduction to enhance speech intelligibility in an audio signal. Described techniques include initially analyzing audio frames in the time domain to identify frames having relatively low power levels. Those frames are then further analyzed in the frequency domain to estimate noise. For example, the initially identified frames may be analyzed at each of multiple frequencies to detect the lowest exhibited power at each of those frequencies. The lowest power values are used as an estimation of noise across the frequency spectrum, and as the basis for calculating a spectral gain for filtering the audio signal in the frequency domain.

Abstract: A fragmentation channelizer has a transmitter channelizing filter bank that fragments a continuous spectral span of a signal input into discrete spectral channels. The discrete spectral channels coincide with available bandwidth segments of a communications channel. A receiver has a receiver filter bank that inputs the discrete spectral channels after transmission over the communications channel and de-fragments the discrete spectral channels into a reconstructed continuous spectral span of the signal input so as to generate a signal output corresponding to the signal input.

Abstract: A system and method is provided for detecting errors in a speech transmission system. A first audio stream is comprised of a plurality of words, upon which a plurality of independent voice-to-text conversions are performed. If it is determined that at least one of the plurality of independent voice-to-text conversions is error free, a text-to-voice conversion of the at least one error-free voice-to-text conversion is performed to create a second audio stream.

Abstract: Method, device and computer program product for processing signals. Signals are received at a plurality of sensors of the device. The initiation of a signal state in which signals of a particular type are received at the plurality of sensors is determined. Responsive to the determining of the initiation of the signal state, data indicating beamformer coefficients to be applied by a beamformer of the device is retrieved from data storage means, wherein the indicated beamformer coefficients are determined so as to be suitable for application to signals received at the sensors in the signal state. The beamformer applies the indicated beamformer coefficients to the signals received at the sensors in the signal state, thereby generating a beamformer output.

Abstract: Disclosed are a method and apparatus for decoding a an audiospeech signal using an adaptive codebook update. The method for decoding speech an audio signal includes: receiving an N+1-th normal frame data that is a normal frame transmitted after an N-th frame that is a loss frame data loss; determining whether an adaptive codebook of a final subframe of the N-th frame is updated or not by using the N-th frame and the N+1-th frame; updating the adaptive codebook of the final subframe of the N-th frame by using a the pitch index of the N+1-the frame; and synthesizing an audio a speech signal of by using the N+1-th frame.

Abstract: A probability model represented as the product of the probability distribution of a mismatch vector g (or clean speech x) with an observed value y as a factor and the probability distribution of a mismatch vector g (or clean speech x) with a confidence index ? for each band as a factor, executes MMSE estimation on the probability model, and estimates a clean speech estimated value x^. As a result, each band influences the result of MMSE estimation, with a degree of contribution in accordance with the level of its confidence. Further, the higher the S/N ratio of observation speech, the more the output value becomes shifted to the observed value. As a result, the output of a front-end is optimized.

Abstract: An audio processing method is disclosed. In the audio processing method, a modified discrete cosine transform (MDCT) algorithm is utilized to transform a present time domain audio signal into a spectrum audio signal. A spreading function (SF) coefficient of each partition domain of the spectrum audio signal is obtained by referencing an SF table, wherein the table is stored in three linear arrays based on non-zero SF-Coefficient values. A masking partitioned energy threshold of each partition domain of the spectrum audio signal is calculated utilizing a logarithmic scale. An audio block type of each partition domain and an SMR of the spectrum audio signal are calculated. Subsequently, the spectrum audio signal is compressed into an audio bit stream according to the audio block type of each partition domain and the SMR. In addition, an audio signal processing apparatus is also disclosed in this invention.

Abstract: A computer implemented method comprising: setting a plurality of frames on a time axis between a first waveform of an input to audio processing and a second waveform of an output from the audio processing, detecting a voice frame and a noise frame in the first and second waveform, calculating a first and second spectrum from the first and second waveform, adjusting level of the first or second spectrum of the noise frame, setting the adjusted first and second spectrum of the noise frame as a third and fourth spectrum, calculating a distortion amount of the noise frame from the third and fourth spectrum, estimating a noise model spectrum from the first or second spectrum, determining a selected frequency by comparison of voice and noise frame spectrum levels, and calculating a distortion amount of the voice frame from the first and second spectrum of the voice frame at the selected frequency.

Abstract: The noise suppression device includes: a shock noise detection unit which receives an input signal including a shock noise and detects a shock noise according to a change of the input signal; and a shock sound suppression unit which receives the shock sound detection result and the input signal so as to suppress the shock sound.

Abstract: In order to compensate tonal changes arising from a multi-path propagation of sound portions during the mixing of multi microphone audio recordings as far as possible it is suggested to form spectral values of respectively overlapping time frames of samples of each a first microphone signal (100) and a second microphone signal (101). The spectral values (300) of the first microphone signal (100) are distributed with formation of spectral values (311) of a first sum signal to the spectral values (301) of a second microphone signal (101) in a first summing level (310), whereat a dynamic correction of the spectral values (300, 301) of one of the two microphone signals (100, 101) occurs. Spectral values (399) of a result signal are formed out of the spectral values (311) of the first sum signal which are subject to an inverse Fourier-transformation and a block junction (FIG. 3).

Abstract: The present invention relates to a method and device for dereverberation of single-channel speech.

Abstract: Disclosed herein are systems, methods, and computer-readable storage devices for processing audio signals. An example system configured to practice the method receives audio at a device to be transmitted to a remote speech processing system. The system analyzes one of noise conditions, need for an enhanced speech quality, and network load to yield an analysis. Based on the analysis, the system determines to bypass user-defined options for enhancing audio for speech processing. Then, based on the analysis, the system can modify an audio transmission parameter used to transmit the audio from the device to the remote speech processing system. The audio transmission parameter can be one of an amount of coding, a chosen codec, an amount of coding, or a number of audio channels, for example.

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: For filtering an audio signal for a non-real-time recipient, a signal module may detect communication of an audio signal to the non-real-time recipient. A management module may filter the audio signal with a high-latency audio filter.

Abstract: The present invention discloses a speech enhancement method and device for mobile phones. By the method and device provided by the present invention, the mobile phone holding state of a user is detected when the user is talking on the phone, so that different denoising solutions will be employed according to the state of the user in holding the mobile phone. When the user holds the mobile phone normally, a solution integrating multi-microphone denoising and single-microphone denoising will be employed to effectively suppress both the steady noise and the non-steady noise; and when the user holds the mobile phone abnormally, a solution of single-microphone denoising will be employed only to suppress the steady noise. The distortion of speech by multi-microphone denoising is avoided, and the speech quality is ensured.

Abstract: An apparatus comprising at least one processor and at least one memory including computer code for one or more programs, the at least one memory and the computer code configured to with the at least one processor to cause the apparatus to at least perform: estimating a signal to noise ratio value for an audio signal; generating a post-filter comprising at least one of: a first formant frequency filter and a second formant frequency filter, wherein the post-filter is dependent on the signal to noise ratio value for the audio signal,

Abstract: A decoder for generating an audio output signal having one or more audio output channels from a downmix signal having one or more downmix channels is provided. The downmix signal encodes one or more audio object signals. The decoder has a threshold determiner for determining a threshold value depending on a signal energy and/or a noise energy of at least one of the of or more audio object signals and/or depending on a signal energy and/or a noise energy of at least one of the one or more downmix channels. Moreover, the decoder has a processing unit for generating the one or more audio output channels from the one or more downmix channels depending on the threshold value.

Abstract: Spatially selective augmentation of a multichannel audio signal is described.

Abstract: Systems and methods for automatic user specific, condition specific communication system intelligibility testing and optimization are provided. The intelligibility of speech for a particular user is determined using a test of intelligibility administered by an interactive voice response (IVR) application running on a communication server. The intelligibility test can be run for a particular user under different conditions. For each user and/or set of conditions, a set of speech signal adjustment parameters can be determined. A set of speech signal adjustment parameters that will enhance the intelligibility of a speech signal for a user are applied when that user is involved in a communication session. The particular set of speech signal adjustment parameters selected can depend on the communication equipment and/or environment associated with the communication session.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for receiving (i) audio data that encodes a spoken natural language query, and (ii) environmental audio data, obtaining a transcription of the spoken natural language query, determining a particular content type associated with one or more keywords in the transcription, providing at least a portion of the environmental audio data to a content recognition engine, and identifying a content item that has been output by the content recognition engine, and that matches the particular content type.

Abstract: Systems, methods and apparatus are described herein for continuously measuring voice clarity and speech intelligibility by evaluating a plurality of telecommunications channels in real time. Voice clarity and speech intelligibility measurements may be formed from chained, configurable DSPs that can be added, subtracted, reordered, or configured to target specific audio features. Voice clarity and speech intelligibility may be enhanced by altering the media in one or more of the plurality of telecommunications channels. Analytics describing the measurements and enhancements may be displayed in reports, or in real time via a dashboard.

Abstract: In prediction of a speech quality evaluation score such as a phone speech, even when a background noise exists, a subjective opinion score is predicted with high precision. A speech quality evaluation system that outputs a predicted value of the subjective opinion score for an evaluation speech such as a far-end speech of a phone, includes a speech distortion calculation unit that conducts, after calculating frequency characteristics of the evaluation speech, a process of subtracting given frequency characteristics from frequency characteristics of the evaluation speech, and calculates the speech distortion on the basis of the frequency characteristics after the subtracting process has been conducted, and a subjective evaluation prediction unit that calculates the predicted value of the subjective opinion score on the basis of the speech distortion.

Abstract: A method of improving signal-to-noise ratio is provided. The method includes acquiring a digital audio signal output from an endec by a host processing unit interface; receiving the digital audio signal output from the host processing unit interface by an audio engine interface; transmitting the digital audio signal from the audio engine interface to an audio engine processor; reducing noise of the digital audio signal by the audio engine processor and then converting the digital audio signal into an analog audio signal to be played.

Abstract: Acoustic noise in an audio signal is reduced by calculating a speech probability presence (SPP) factor using minimum mean square error (MMSE). The SPP factor, which has a value typically ranging between zero and one, is modified or warped responsive to a value obtained from the evaluation of a sigmoid function, the shape of which is determined by a signal-to-noise ratio (SNR), which is obtained by an evaluation of the signal energy and noise energy output from a microphone over time. The shape and aggressiveness of the sigmoid function is determined using an extrinsically-determined SNR, not determined by the MMSE determination. The extrinsically-determined SNR is obtained from a long term history of previously-determined speech presence probabilities and a long term history of previously-determined noise histories.

Abstract: Acoustic noise in an audio signal is reduced by calculating a speech probability presence (SPP) factor using minimum mean square error (MMSE). The SPP factor, which has a value typically ranging between zero and one, is modified or warped responsive to a value obtained from the evaluation of a sigmoid function, the shape of which is determined by a signal-to-noise ratio (SNR), which is obtained by an evaluation of the signal energy and noise energy output from a microphone over time. The shape and aggressiveness of the sigmoid function is determined using an extrinsically-determined SNR, not determined by the MMSE determination.

Abstract: Acoustic noise in an audio signal is reduced by calculating a speech probability presence (SPP) factor using minimum mean square error (MMSE). The SPP factor, which has a value typically ranging between zero and one, is modified or warped responsive to a value obtained from the evaluation of a sigmoid function, the shape of which is determined by a signal-to-noise ratio (SNR), which is obtained by an evaluation of the signal energy and noise energy output from a microphone over time.

Abstract: The present invention relates to means and methods of automated difference recognition between speech and music signals in voice communication systems, devices, telephones, and methods, and more specifically, to systems, devices, and methods that automate control when either speech or music is detected over communication links. The present invention provides a novel system and method for monitoring the audio signal, analyze selected audio signal components, compare the results of analysis with a pre-determined threshold value, and classify the audio signal either as speech or music.

Abstract: Methods and systems for using pitch predictors in speech/audio coders are provided. Techniques for optimal pre- and post-filtering are presented, and a general result that post-filtering is more effective than pre-filtering is derived. A practical paired-zero filter design for the low-rate regime is proposed, and this design is extended to handle frequency-dependent periodicity levels. Further, the methods described provide a general performance measure for a post-filter that only uses information available at the decoder, thereby allowing for the optimization or selection of a post-filter without increasing the rate.

Abstract: A device and method are disclosed for testing the intelligibility of audio announcement systems. The device may include a microphone, a translation engine, a processor, a memory associated with the processor, and a display. The microphone of the analyzer may be coupled to the translation engine, which in-turn may be coupled to the processor, which is in-turn may be coupled to the memory and the display. The translation engine can convert audio speech input from the microphone into data output. The processor can receive the data output and can apply a scoring algorithm thereto. The algorithm can compare the received data against data that is stored in the memory of the analyzer and calculates the accuracy of the received data. The algorithm may translate the calculated accuracy into a standardized STI intelligibility score that is then presented on the display of the analyzer.

Abstract: In a pulse encoding and decoding method and a pulse codec, more than two tracks are jointly encoded, so that free codebook space in the situation of single track encoding can be combined during joint encoding to become code bits that may be saved. Furthermore, a pulse that is on each track and required to be encoded is combined according to positions, and the number of positions having pulses, distribution of the positions that have pulses on the track, and the number of pulses on each position that has a pulse are encoded separately, so as to avoid separate encoding performed on multiple pulses of a same position, thereby further saving code bits.

Abstract: MDCT or FFT-based audio coding algorithms often have the problem named here spectral pre-echoes when coding an energy attack signal. This invention presents several possibilities to avoid the spectral pre-echoes existing in decoded signal segment before the energy attack point. The spectral envelope before the attack point can be improved by performing spectrum smoothing, replacing the segment of having spectral pre-echoes or filtering the segment with a combined filter obtained by doing LPC analysis.

Abstract: A method, system and apparatus are shown for identifying non-language speech sounds in a speech or audio signal. An audio signal is segmented and feature vectors are extracted from the segments of the audio signal. The segment is classified using a hidden Markov model (HMM) that has been trained on sequences of these feature vectors. Post-processing components can be utilized to enhance classification. An embodiment is described in which the hidden Markov model is used to classify a segment as a language speech sound or one of a variety of non-language speech sounds. Another embodiment is described in which the hidden Markov model is trained using discriminative learning.