Abstract: A method, system, and computer-program product includes identifying a set of heterogeneous sensors, configuring a plurality of model training compositions for each of the set of heterogeneous sensors, computing, for each of the plurality of model training compositions, a first efficacy metric value based on predictive outputs of the at least two machine learning models, identifying, for each sensor of the set of heterogeneous sensors, a champion model training composition of the subject sensor, the champion model training composition having a highest efficacy metric value, and electing, from a plurality of champion model training compositions corresponding to the champion model training compositions identified for each sensor of the set of heterogeneous sensors, an overall champion model training composition corresponding to a champion sensor of the set of heterogeneous sensors based on an assessment of second efficacy metric values of the plurality of champion model training compositions.

Abstract: An apparatus for estimating bio-information includes a bio-signal obtainer configured to obtain a bio-signal from an object; and a processor configured to: obtain function values by applying a predetermined function to sections of the bio-signal, corresponding to respective windows of a predetermined size, while sliding a window on a time axis of the bio-signal, obtain an oscillometric waveform envelope based on the obtained function values, and estimate bio-information of the object based on the oscillometric waveform envelope.

Abstract: An ultrasound imaging system includes an array of ultrasound transducer elements chat send ultrasound energy into an object when energized for respective transmission time periods and provide responses to ultrasound energy emitted from the object for respective reception time periods, a reception modulation circuit modulating the responses with irregular sequences of modulation coefficients, a combiner circuit combining the modulated responses, and an image reconstruction processor configured to computer-process the combined modulated responses into one or more images of the object.

Abstract: A physical layer transceiver for a wireline communications system includes a receive path for communicating signals received from the medium to a functional circuit, a transmit path for communicating signals received from the functional circuit onto the medium, and echo cancellation circuitry for cancelling from the receive path echoes of signals transmitted on the transmit path. The echo cancellation circuitry includes a plurality of high-dynamic-range filter segments, a plurality of low-dynamic-range filter segments, and control circuitry configured to detect on the receive path the echoes of the signals transmitted on the transmit path, classify each echo signal as having a low dynamic range or a high dynamic range, and deploy a high-dynamic-range segment to a location of an echo signal having a high dynamic range, and deploy a low-dynamic-range segment to a location of an echo signal having a low dynamic range.

Abstract: Hybrid audio beamforming systems and methods with narrower beams and improved directivity are provided. The hybrid audio beamforming system includes a time domain beamformer for processing upper frequency band signals of an audio signal using a time domain beamforming technique, and a frequency domain beamformer for processing groups of lower frequency band signals of the audio signal using frequency domain beamforming techniques.

Abstract: An adaptive processor implements partial updates when it adjusts weights to optimize adaptation criteria in signal estimation, parameter estimation, or data dimensionality reduction algorithms. The adaptive processor designates some of the weights to be update weights and the other weights to be held weights. Unconstrained updates are performed on the update weights, whereas updates to the set of held weights are performed within a reduced-dimensionality subspace. Updates to the held weights and the update weights employ adapt-path operations for tuning the adaptive processor to process signal data during or after tuning.

Abstract: A digital filter for filtering a pulse density modulation (PDM) signal is presented. The filter has a first filter circuit to receive an input signal and to provide a filtered input signal at successive time steps which include a first filtered value at the first time step and a second filtered value at a second time step. The filter also has a quantizer to provide an output signal comprising output values at successive time steps and a filter variable circuit with a first multiplication circuit to receive the first filter variable, and divide the first filter variable by a first gain factor and a first summing circuit configured to receive the divided first filter variable, receive the output signal, and add the divided first filter variable and the first output value and a second multiplication circuit and a delay circuit.

Abstract: An acoustic vector sensor has an array of sensors to detect at least the bearing of a target. The acoustic vector sensor or hydrophone with sensor array avoids the need to deploy multiple hydrophones each with a single sensor. The array of sensor signals can be processed using any one of a number of methods.

Abstract: Techniques related to a low-power, low-memory multiply and accumulate (MAC) unit are described. In an example, the MAC unit performs a MAC operation that represents a multiplication of numbers. At least the bit representation of one number is compressed based on a quantization and a clustering of quantization values, whereby index bits are used instead of the actual bit representation. The index bits are loaded in an index buffer and a bit representation of another number is loaded in an input buffer. The index bits are used in a lookup to determine whether the corresponding bit representation and shift operations are applied to the input buffer based on this bit representation, followed by accumulation operations.

Abstract: A computer-implemented method includes applying a filter to input data based on an initial set of parameters to generate an initial feature map. The filter is configured to activate a filter function that involves a periodic function. The method includes performing a first linear transform on the initial feature map based on a subset of a first set of parameters to generate a first linear transform. The method includes applying the filter to the input data based on another subset of the first set of parameters to generate a first feature map. The method includes performing a multiplicative operation on the first linear transform and the first feature map to generate a first product. The method includes performing a second linear transform on the first product based on a subset of a second set of parameters to generate a second linear transform. The method includes generating output data that takes into account at least the second linear transform.

Abstract: Disclosed are a speech processing method and a speech processing apparatus in a 5G communication environment through speech processing by executing embedded artificial intelligence (AI) algorithms and/or machine learning algorithms. The speech processing method includes determining a temporary pause of reception of a first spoken utterance, outputting a first spoken response utterance as a result of speech recognition processing of a second spoken utterance received after the temporary pause, determining, as an extension of the first spoken utterance, a third spoken utterance that is received after outputting the first spoken response utterance, deleting, using a deep neural network model, a duplicate utterance part from a fourth spoken utterance that is obtained by combining the first and the third spoken utterance, and outputting a second spoken response utterance as a result of speech recognition processing of the fourth spoken utterance from which the duplicate utterance part has been deleted.

Abstract: A loudspeaker device includes at least one loudspeaker, a loudspeaker holder holding the at least one loudspeaker in a reference range away from the ear of a user by a reference distance, a first microphone collecting an environmental sound and outputting an electrical signal, a second microphone attached to a position where a sound output from the at least one loudspeaker is collected, the second microphone collecting a synthetic sound synthesized from the sound output from the at least one loudspeaker and the environmental sound and outputting an electrical signal, and a processor controlling the at least one loudspeaker so as to output a sound for reducing the environmental sound based on the electrical signals representing the sounds collected by the first microphone and the second microphone.

Abstract: Embodiments of the disclosure implement an application of an adaptive filter bank that is used to characterize the heat transfer of a volume in a thermal system, to estimate temperature and power consumption, and to improve performance characteristics in applications including optimal temperature control and diagnostics. In some embodiments, the adaptive filter bank is an iterative solution, comprised of a collection of adaptive filters defined to consume incident signals, produce an aggregate reference signal, estimate an error relative to an observed primary signal, and modify thermal coefficients to converge on a solution. For example, the incident signals are comprised of properties related to active, passive, solar irradiance, and unobserved heat transfer. A reference signal is an estimate of a primary signal, related to the rate of heat transfer or temperature change. Thereupon, the thermal coefficients are modified in an adaptive process to include gradient descent, which minimizes estimation error.

Abstract: According to the present disclosure, an additional sound generating unit detects, as a noise frequency, a frequency of a noise at a control point and generates an additional sound signal including signal components with additional frequencies different from the noise frequency. A canceling signal generating unit generates a canceling signal that cancels the noise at the control point. An emission unit outputs a control sound signal, generated by adding the additional sound signal to the canceling signal, to a loudspeaker and makes the loudspeaker emit the control sound.

Abstract: Various embodiments provide methods, apparatuses, systems, or computer program products for providing a signal with dynamically shaped noise. In an example embodiment, the system comprises a dynamic filter, wherein a dynamic filter is a filter that is capable of switching between at least two responses; a signal generator configured to generate a signal; and a controller configured to control operation of the signal generator and selection of an operating response from the at least two responses of the dynamic filter. The controller causes the signal generator to generate a signal, the signal is provided to the dynamic filter, and the controller causes the dynamic filter to filter the signal in accordance with the operating response. The filtered signal is then provided to an electrical component of the system to cause at least a portion of the system to perform a function.

Abstract: An endpoint receives audio from a remote endpoint. A first signal corresponding to the audio is received at an adaptive filter, and a filtered signal is generated. First audio is emitted at a loudspeaker based on the first signal. A microphone collects second audio which is based on the first audio. The microphone signal emits a signal based on the second audio. The filtered signal is subtracted from microphone signal to generate an adapted signal, the adapted signal having an energy level. The adapted signal is then transmitted to a double-talk detector, the double-talk detector configured to allow transmission of the adapted signal to the remote endpoint when the energy level of the adapted signal exceeds an energy threshold. The degree of cross-correlation between the first signal and the adapted signal is determined (iteratively). If the cross-correlation exceeds a cross-correlation threshold, the energy threshold of the double-talk detector is raised.

Abstract: A method is provided for a correction of an audio signal (AS) which is formed using an electroacoustic transducer. The method comprises producing a corrected audio signal (CAS) based on the audio signal (AS) according to the formal I: in which the coefficient B amounts from ?10 to 0 and the coefficient C amounts from 0 to 0.04.

Abstract: The present disclosure relates to a method for processing input visual data using a generated algorithm based upon input visual data and the output of a calculated energy function. According to a first aspect of the disclosure, there is provided a method for enhancing input visual data using an algorithm, the method comprising evaluating gradients of the output of an energy function with respect to the input visual data; using the gradient output to enhance the input visual data; and outputting the enhanced visual data.

Abstract: Provided is a target speech signal extraction method for robust speech recognition including: (a) receiving information on a direction of arrival of the target speech source with respect to the microphones; (b) generating a nullformer by using the information on the direction of arrival of the target speech source to remove the target speech signal from the input signals and to estimate noise; (c) setting a real output of the target speech source using an adaptive vector w(k) as a first channel and setting a dummy output by the nullformer as a remaining channel; (d) setting a cost function for minimizing dependency between the real output of the target speech source and the dummy output using the nullformer by performing independent component analysis (ICA); and (e) estimating the target speech signal by using the cost function, thereby extracting the target speech signal from the input signals.

Abstract: A method of determining a hybrid index may include obtaining, by a computer device, financial transaction information about two or more financial products over a duration. The computer device may be configured to filter the financial transaction information to produce enhanced financial transaction information. The computer device may then determine a financial index value using the financial transaction information and the enhanced financial transaction information.

Abstract: The present embodiments include methods of time-frequency analyzing signals. Some embodiments provide methods of processing signals comprising: adaptively partitioning at least a portion of a signal in a time domain into a plurality of segments of the signal; and transforming each of the segments of the signal producing respective expansions in a frequency domain and obtaining respective samples of the windows of signal in the frequency domain.

Abstract: A partitioned block frequency domain adaptive filter device includes a frequency domain adaptive filter configured for filtering a frequency domain representation of a time domain input signal depending on a set of filter coefficients consisting of a plurality of blocks of filter coefficients in order to produce a filtered signal; a plurality of parallel arranged filter update blocks; wherein each of the filter update blocks includes an adaptation module configured for executing an adaptation sequence including the steps of calculating an approximation of a constrained gradient update for the filter coefficients of the respective block of filter coefficients, and calculating a cumulative error introduced on the unconstrained gradient update; wherein each of the filter update blocks includes a correction module configured for executing a correction sequence including the steps of calculating a corrected constrained gradient update for the filter coefficients of the respective block of filter coefficients.

Abstract: A method for processing data includes receiving a depth map of a scene containing at least a humanoid head, the depth map comprising a matrix of pixels having respective pixel depth values. A digital processor extracts from the depth map a curvature map of the scene. The curvature map includes respective curvature values of at least some of the pixels in the matrix. The curvature values are processed in order to identify a face in the scene.

Abstract: A method and apparatus for operating an internal combustion engine is disclosed. A signal generated by the oxygen concentration sensor is converted into a first signal indicative of an air/fuel ratio in the engine cylinder. A second signal indicative of an expected air/fuel ratio in the engine cylinder due to a fuel injection is generated and filtered to obtain a filtered signal and used to operate the engine. The filtered signal is obtained by sampling values of the first and second signals. A time constant is calculated based on the sampled values of the first and second signals and a value of the first signal sampled during a preceding control cycle. A value of the filtered signal is calculated based on the calculated time constant, the sampled value of the second signal and a value of the filtered signal calculated during the preceding control cycle.

Abstract: A method performs frequency-dependent reduction of noise in an input signal, wherein the input signal is allocated to a main signal path and to an auxiliary signal path. A filter bank separates the input signal into a plurality of frequency bands each having a certain bandwidth and center frequency. In the auxiliary signal path, the bandwidth of the signal component is reduced in each frequency band thus forming a reduced frequency band in each case. Wherein in each reduced frequency band, a reduction parameter is determined from the signal component of the input signal, and wherein in the main signal path in each frequency band, noise in the input signal is reduced by use of the reduction parameter in that frequency band.

Abstract: In accordance with an embodiment the method includes temporarily configuring the vector processor with a new set of vectoring coefficients during one or more selected symbol positions; restoring the current set of vectoring coefficients outside the one or more selected symbol positions; obtaining at least one error measure over respectively at least one line of the group of vectored lines during the one or more selected symbol positions; and determining a suitability indication for the new set of vectoring coefficients based on the obtained at least one error measure.

Abstract: Compression of multidimensional sonic data is disclosed. Example methods disclosed herein to compress input multidimensional sonic data include decorrelating the input multidimensional sonic data to determine decorrelated multidimensional sonic data. In some such examples, the input multidimensional sonic data has a plurality of dimensions corresponding to a respective plurality of azimuthal directions from which the input multidimensional sonic data was obtained. In such examples, the decorrelated multidimensional sonic data is decorrelated among the plurality of dimensions of the input multidimensional sonic data. Such example methods also include processing the decorrelated multidimensional sonic data to determine compressed data representative of the input multidimensional sonic data.

Abstract: The adaptive filter for sparse system identification is an adaptive filter that uses an algorithm in the feedback loop that is designed to provide better performance when the unknown system model is sparse, i.e., when the filter has only a few non-zero coefficients, such as digital TV transmission channels and echo paths. The algorithm is a least mean square algorithm with filter coefficients updated at each iteration, as well as a step size that is also updated at each iteration. The adaptive filter may be implemented on a digital signal processor (DSP), an application-specific integrated circuit (ASIC), or by field-programmable gate arrays (FPGAs).

Abstract: A multi-channel acoustic echo cancellation (AEC) system that includes a step-size controller that dynamically determines a step-size value for each channel and each tone index on a frame-by-frame basis. The system determines the step-size value based on a normalized squared cross-correlation (NSCC) between an estimated echo signal and an error signal, allowing the AEC system to converge quickly when an acoustic room response changes while providing stable steady-state error by avoiding misadjustments due to noise sensitivity and/or near-end speech. The step-size value can be determined using fractional weighting that takes into account a signal strength of each channel.

Abstract: A digital low pass filter for producing an output value given a target value includes a memory which stores a scaling factor, a previous output value, a previous intermediate value, and the target value; the difference between the target value and the previous output value is evaluated, and then multiplied by the scaling factor to produce an intermediate value; the previous intermediate value is multiplied by the scaling factor minus unity; the output value is evaluated by summing the previous output value, twice the intermediate value, and the previous intermediate value multiplied by the scaling factor minus unity; the output value is then stored in memory as the previous output value, and the intermediate value as the previous intermediate value, such that the filter provides a second-order response but requires fewer hardware multipliers than the direct form implementation of a second-order filter.

Abstract: A receiver includes a static I/Q calibration block and a correlation/integration block. The static I/Q calibration block is configured to substantially eliminate mismatches between in-phase and quadrature components of a portion of the spectrum having associated I/Q mismatches that are relatively frequency-independent. The correlation/integration block is configured to substantially eliminate mismatches between the in-phase and quadrature components of portions of the spectrum having associated I/Q mismatches that are relatively frequency-dependent in accordance with a pair of signals generated by the static I/C calibration block.

Abstract: A method for speeding up a filter tracking speed includes scaling filter coefficients based at least in part on an uplink/downlink configuration in a time division duplex (TDD) or a multimedia broadcast single frequency network (MBSFN) system. The method also includes applying scaled filter coefficients during at least one downlink subframe to control a filter tracking speed.

Abstract: The present inventions are related to systems and methods for data processing, and more particularly to systems and methods for data detection. As one example, a data processing system is described that includes a variance calculation circuit operable to calculate a variance of a data input; and a branch metric calculation circuit operable to calculate a branch metric based at least in part on the variance.

Abstract: A method and apparatus of suppressing a vocoder noise are provided. The method includes receiving from a channel decoder a vocoder frame and first information, the first information indicating whether the vocoder frame has an error, generating speech data by performing voice decoding on the vocoder frame, determining whether a tonal noise has been detected in the speech data, if the first information indicates that the vocoder frame has an error, and attenuating the volume of the speech data and outputting the volume-attenuated speech data through a speaker, upon detection of the tonal noise in the speech data.

Abstract: An electronic receiver may comprise nonlinear distortion modeling circuitry, interference estimation circuitry, and sequence estimation circuitry. The receiver may receive an orthogonal frequency division multiplexing (OFDM) symbol in the form of an electromagnetic signal. The nonlinear distortion modeling circuitry may generate a nonlinear distortion model that models nonlinear distortion introduced to the received electromagnetic signal en route to the sequence estimation circuitry. The interference estimation circuitry may estimate inter-subcarrier interference present in the received OFDM symbol based on the generated nonlinear distortion model. The estimating of the inter-subcarrier interference may comprise applying the nonlinear distortion model to one or more candidate vectors generated by the sequence estimation circuitry. The sequence estimation circuitry may sequentially process a plurality of received virtual subcarrier values of the OFDM symbol using the estimated inter-subcarrier interference.

Abstract: A multi-stage adaptive filter is disclosed, which exhibits a smaller mean square error than in prior art adaptive filters. The adaptive filter selectively manipulates the weights, in multiple stages, so as to achieve a global minimum of the error function, such that the filtered signal has as small an error as possible.

Abstract: A transmission method and system for use in a base station of a mobile communication network. It provides a simple and efficient technique which improves the performance of prior art solutions for MIMO user equipments (especially in bad or medium radio conditions scenarios).

Abstract: Various embodiments of the present invention provide systems and methods for calibrating a data processing circuit. For example, a method for calibrating a data processing circuit is discussed that includes providing a digital filter, providing a detector circuit, and providing an analog filter. Operation of the digital filter is at least in part governed by filter taps that correspond to a filter tap constraint value. Operation of the detector circuit is at least in part governed by a target parameter. Operation of the analog filter is at least in part governed by an analog parameter that is one of a plurality of analog parameters. The methods further include selecting a target parameter, and calculating the filter tap constraint value based on the target parameter. Combinations of the target parameter, the calculated filter tap constraint value, and each of the plurality of analog parameters are applied to identify the analog parameter.

Abstract: An event-driven tracking analog to digital converter (ADC) architecture is proposed. The proposed architecture has less sensitivity to amplifier and DAC non-linearity, reduces the swing and dynamic common-mode range requirement of the operational transconductance amplifier (OTA) and comparators, respectively.

Abstract: Systems and methods for noise suppression using noise subtraction processing are provided. The noise subtraction processing comprises receiving at least a primary and a secondary acoustic signal. A desired signal component may be calculated and subtracted from the secondary acoustic signal to obtain a noise component signal. A determination may be made of a reference energy ratio and a prediction energy ratio. A determination may be made as to whether to adjust the noise component signal based partially on the reference energy ratio and partially on the prediction energy ratio. The noise component signal may be adjusted or frozen based on the determination. The noise component signal may then be removed from the primary acoustic signal to generate a noise subtracted signal which may be outputted.

Abstract: This invention relates to a color controlled light source comprising a plurality of colored light elements, and a plurality of (filtered) photo detectors, having different spectral characteristics covering all or most of the total spectrum of the light elements. The (filtered) photo detectors detect the light output of the light source, and generate corresponding detection signals. The light source further has a color control unit for generating driving signals to the light elements on the basis of the detection signals and a predetermined target color point of the light output of the light source, and a modulator for individual signature modulation of the driving signal to each one of said light elements. A corresponding demodulator is provided for demodulation of the detection signals and extraction, from each detection signal, of actual values of the light outputs of the light elements.

Abstract: A method for performing orthogonal frequency division multiplexing (OFDM)/offset quantization amplitude modulation (OQAM) includes obtaining a data burst. The method includes performing weighted circularly convolved filtering modulation on the data burst to produce an output signal. The method further includes a first wireless device transmitting the output signal to a second wireless device. The second wireless device receives an input signal from the first wireless device, and the second wireless devices performs weighted circularly convolved demodulation filtering on the input signal to produce the data burst.

Abstract: Methods, systems, and computer readable media can be operable to dynamically determine an optimal number of equalizer taps to be used at an upstream receiver based upon existing conditions on an upstream channel. Pre-equalizer coefficients associated with an upstream channel can be used to determine an optimal number of equalizer taps to be used at an upstream receiver. The number of equalizer taps can be dynamically determined for an individual customer premise equipment device or for a plurality of customer premise equipment devices using the same upstream channel.

Abstract: An adaptive digital filtering method and apparatus for touch data in a touch system is described. Adaptive digital filtering involves estimating a signal-to-noise ratio (SNR) for stored touch data, and reducing a digital filter order if the SNR is high and increasing the digital filter order if the SNR is small.

Abstract: A multi-core processor is used in a portable device that has first and second image sensors spaced from each other for capturing images of a scene from slightly different perspectives. The multi-core processor has a first image sensor interface for receiving data from the image sensor, a second image sensor interface for receiving data from the second image sensor, multiple processing units and, the four processing units and the first and second sensor interfaces being integrated onto a single chip. The processing units are configured to simultaneously process the data from the first and second image interfaces to generate stereoscopic image data.

Abstract: A tunable receiver system uses programmable notch filters to identify available channel pairs for transmitting and receiving data via a femtocell base station. In addition, one of the programmable notch filters may be used to suppress infiltration of the transmit path signal into the receiver path of the receiver device. The other programmable notch filter may be used to suppress a blocker signal identified by the receiver device.

Abstract: A multi-core processor is used in a portable device that has first and second image sensors spaced from each other for capturing images of a scene from slightly different perspectives. The multi-core processor has a first image sensor interface for receiving data from the image sensor, a second image sensor interface for receiving data from the second image sensor, multiple processing units and, the four processing units and the first and second sensor interfaces being integrated onto a single chip. The processing units are configured to simultaneously process the data from the first and second image interfaces to generate stereoscopic image data.

Abstract: A method and apparatus for communication in a system employing differing transmission protocols. It is determined whether a first time slot of one transmission protocol is adjacent to a second time slot of another transmission protocol, and if a first symbol within the first time slot is adjacent to a second symbol of the second time slot. A pilot power and/or a number of pilot sub-carriers or tones of the first symbol is increased and channel estimation is performed on the first symbol.

Abstract: The present technique relates to a reception device and a reception method which can improve equalization performance. An equalization processing unit has a time domain equalization unit which equalizes a received signal in a time domain and a frequency domain equalization unit which is provided in parallel to the time domain equalization unit and which equalizes the received signal in a frequency domain, and performs control of switching between the time domain equalization unit and the frequency domain equalization unit. The present technique can be applied to, for example, equalize a signal of data transmitted by way of single carrier transmission or data transmitted by way of multicarrier transmission.

Abstract: Provided is a system for generating coefficient values. The system may include a base function generator and a series of accumulators including a leading and a last accumulator. In the series of accumulators, the data output of each accumulator, except the last, may be coupled to the data input of a successive adjacent accumulator. The base function generator may be configured to output, to the leading accumulator, a series of data values that may correspond to a base function that is a specified order derivative of a filter function. Each accumulator may be configured to: add a data value currently at its data input to a currently stored data value to produce an updated data value that may correspond to a respective value of a specified order integral of the base function; store the updated data value in the accumulator; and output the updated data value at its data output.