44 results about "Spectral envelope" patented technology

The invention relates to the technical field of sleep quality evaluation, in particular to a sleep monitoring optimization system based on artificial intelligence. The system comprises the steps of collecting sleep monitoring data of a user at different collection moments, extracting dominant frequency components in snore audio data through spectrum envelope analysis, extracting audio primitives based on the dominant frequency components, inputting the audio primitives into a snore classification model, outputting a snore category corresponding to the snore audio data of the user, and outputting the snore category corresponding to the snore audio data of the user. Carrying out Bayesian inversion analysis on the brain wave data to obtain a posterior sample of random field model parameters, executing reliability analysis based on the posterior sample to calculate a posterior failure probability, obtaining a first sleep quality index of the user based on the posterior failure probability, obtaining a second sleep quality index of the user based on vital sign data analysis, and sending the second sleep quality index to the user; and obtaining a user sleep quality evaluation result based on the snore type, the user first sleep quality index and the user second sleep quality index. The accuracy and efficiency of user sleep quality evaluation can be improved.

The present invention discloses a continuous speech formant extraction method based on peak selection, comprising: performing a preprocessing operation on a single frame of input speech; using a linear prediction method to preliminarily estimate the peak value in the spectral envelope of the speech frame; establishing a reference point and a formant trough, and then using a peak selection method to establish a mapping relationship between the peak value and the reference point; using the mapping relationship between the peak value and the reference point and the formant trough to determine the formant of the speech frame; and performing formant estimation on the continuous speech: dividing the continuous speech into frames according to different frame numbers, using the above algorithm to loop 100 times to obtain the formant parameters under different frame number tests, averaging the results after 100 loops, and obtaining the final result after smoothing. The method of the present invention can eliminate the influence of merged peaks and false peaks, and has a fast convergence speed and strong robustness.

The invention discloses a micro-seismic signal classification method based on genetic algorithm optimization. The method comprises the following steps: data acquisition: acquiring waveform data containing multiple types of micro-seismic event forms; preprocessing: carrying out mean value removal and band-pass filtering preprocessing on the obtained waveform data, and removing baseline drift and high-frequency noise interference; envelope calculation: respectively calculating a time domain envelope and a frequency spectrum envelope for the preprocessed microseismic signal; feature extraction: extracting a plurality of statistics and time frequency parameters from the time domain envelope and the frequency spectrum envelope to form a feature complete set; feature selection and classifier optimization: utilizing a genetic algorithm to perform joint search optimization on structure parameters and feature subsets of the multi-layer perceptron; and training and verification: training the multilayer perceptron by using the optimized feature subset and the network structure, and evaluating the classification accuracy on a verification set. The method can realize high-precision classification, has robustness to noise, and can be widely applied to the fields of microseism monitoring, blasting event monitoring, other vibration source identification and the like.

The application discloses a bone conduction speech conversion method based on spectral envelope mapping, comprising the following steps: pre-processing and linear prediction analysis of the bone conduction speech signal, and calculating the LP filter coefficient; mapping the LSF coefficient of the air conduction speech signal corresponding to the bone conduction speech signal by using the trained neural network; controlling the minimum error of the original bone conduction speech signal and the synthesized air conduction speech signal according to the timbre weighting characteristic of the bone conduction speech characteristic; estimating the integer pitch of the bone conduction speech signal through the timbre weighting filter, obtaining the reference signal by passing the linear prediction residual signal through the timbre weighting filter, estimating the fractional pitch to obtain the adaptive codebook vector; obtaining the new reference signal by subtracting the adaptive codebook vector from the reference signal, searching for the optimal excitation in the fixed codebook; synthesizing the air conduction speech signal by combining the optimal excitation and the LP filter of the air conduction speech signal, and correcting the air conduction speech signal.

An apparatus for generating a bandwidth extended signal from an input signal includes a patch generator and a combiner. The input signal is represented for first and second bands by first and second resolution data, respectively, the second resolution being lower than the first. The patch generator generates first and second patches from the first band of the input signal according to first and second patching algorithms, respectively. A spectral density of the second patch generated using the second patching algorithm is higher than a spectral density of a first patch generated using the first patching algorithm. The combiner combines both patches and the first band of the input signal to obtain the bandwidth extended signal. The apparatus scales the input signal according to the first and second patching algorithms or scales the first and second patches, so that the bandwidth extended signal fulfills a spectral envelope criterion.

An improved concept for coding sample values of a spectral envelope is obtained by combining spectrotemporal prediction on the one hand and context-based entropy coding the residuals, on the other hand, while particularly determining the context for a current sample value dependent on a measure of a deviation between a pair of already coded / decoded sample values of the spectral envelope in a spectrotemporal neighborhood of the current sample value. The combination of the spectrotemporal prediction on the one hand and the context-based entropy coding of the prediction residuals with selecting the context depending on the deviation measure on the other hand harmonizes with the nature of spectral envelopes.

An apparatus for generating a bandwidth extended signal from an input signal includes a patch generator and a combiner. The input signal is represented for first and second bands by first and second resolution data, respectively, the second resolution being lower than the first. The patch generator generates first and second patches from the first band of the input signal according to first and second patching algorithms, respectively. A spectral density of the second patch generated using the second patching algorithm is higher than a spectral density of a first patch generated using the first patching algorithm. The combiner combines both patches and the first band of the input signal to obtain the bandwidth extended signal. The apparatus scales the input signal according to the first and second patching algorithms or scales the first and second patches, so that the bandwidth extended signal fulfills a spectral envelope criterion.

This invention discloses a cross-receiver radio frequency fingerprinting method, system, and terminal device, relating to the field of wireless communication security technology. The invention estimates the carrier frequency offset of the received IQ signal by decoupling the transmitter and receiver carrier frequency offsets; extracts the slope and curvature of the frequency domain phase as domain-invariant features; constrains the consistency of features of the same device across different receivers based on contrastive learning; employs optimal transmission alignment of the spectral envelope distribution of different receivers to reduce inter-domain offset; and achieves cross-receiver device identification through joint optimization of invariant features. By estimating the carrier frequency offset and decoupling the transmitter and receiver frequency offsets, extracting the frequency domain phase slope and curvature as domain-invariant features, and combining contrastive learning to constrain the consistency of features of the same device across receivers with optimal transmission alignment of the spectral envelope distribution of different receivers, this invention achieves reliable device access authentication and identification under receiver-less conditions, improving the model's generalization ability and stability.

An apparatus for generating a bandwidth extended signal from an input signal includes a patch generator and a combiner. The input signal is represented for first and second bands by first and second resolution data, respectively, the second resolution being lower than the first. The patch generator generates first and second patches from the first band of the input signal according to first and second patching algorithms, respectively. A spectral density of the second patch generated using the second patching algorithm is higher than a spectral density of a first patch generated using the first patching algorithm. The combiner combines both patches and the first band of the input signal to obtain the bandwidth extended signal. The apparatus scales the input signal according to the first and second patching algorithms or scales the first and second patches, so that the bandwidth extended signal fulfills a spectral envelope criterion.

The invention is applicable to the technical field of optical biosensors, and provides an optical biosensor with a single ring-double ring mode switching function, which comprises a broadband light source, an optical signal is transmitted to an optical switch I, an optical sensing chip is respectively connected with the optical switch I and an optical switch II through an optical fiber array, and the optical switch II is connected with a spectrum analyzer. The optical fiber array is connected with the optical sensing chip and the optical switch and used for controlling switching and detection of a single-ring mode or a double-ring mode, when liquid to be detected causes high absorption of the sensing ring, the single-ring sensing mode is adopted, otherwise, the double-ring sensing mode is adopted, and when double-ring sensing is adopted, the single-ring sensing mode is adopted. The reference ring is tuned at the two ends of the thermode of the reference ring, the effective refractive index of the reference ring is changed, when the change of the refractive index is small, characterization of a to-be-measured substance is achieved through the spectral envelope drift distance, when the change of the refractive index is large, characterization of the to-be-measured substance is conducted through the change of FSR, and high-sensitivity and large-range sensing is achieved.

The systems and methods disclosed herein generate a personalized inner voice audio output that replicates or otherwise shares the acoustic characteristics of a speaker's self-perceived voice, which can differ from their externally perceived voice due to bone conduction effects. The systems and methods disclosed herein can generate a provisional voice clone (e.g., parameters of a voice model) of the speaker's voice (e.g., using a recording of the speaker's voice), and can apply a frequency shift to compensate for the absence of bone-conducted low-frequency adjustment that occurs during natural speech production. A trained artificial intelligence model predicts and applies one or more additional acoustic parameter adjustment values (e.g., formant structure, spectral envelope, prosodic patterns) to the provisional voice clone based on one or more factors (e.g., demographic, anatomical, content, environment). The provisional voice clone can be iteratively refined based on received user feedback (e.g., until the output aligns with the speaker's perception of their inner voice).

Systems and techniques are provided for processing audio data. For example, a process can include generating a first sub-vector of audio features corresponding to a first combination of spectral envelope features of the audio, energy features of the audio, and pitch features of the audio, and generating a second sub-vector of audio features corresponding to a second combination of the spectral envelope features, energy features, and pitch features. A first neural network-based autoencoder can be used to generate a first encoded representation corresponding to the first sub-vector of audio features. A second neural network-based autoencoder can be used to generate a second encoded representation corresponding to the second sub-vector of audio features. The first encoded representation and the second encoded representation can be transmitted to a decoder.

The invention relates to the technical field of voice generation, in particular to a personification voice generation method and system for a carry-on AI partner, and the method comprises the following steps: collecting an external environment sound simulation signal through a carry-on equipment microphone array, obtaining an environment acoustic harmonic feature set, storing the environment acoustic harmonic feature set, and obtaining a voice simulation signal; and obtaining environment frequency spectrum characteristic data. According to the method, after a sound channel regular curled spectrum sequence is subjected to time domain smoothing to form frequency domain curled modulation data, a high-frequency-band spectrum envelope interval is extracted, amplitude gain improvement is executed according to a lip radiation impedance coefficient, then the frequency domain curled modulation data is superposed, and anthropomorphic interaction voice signals are generated. And high-frequency friction and gingival blasting details are enhanced without damaging the overall spectrum continuity, the environmental suitability, the anti-interference stability, the formant controllability and the high-frequency detail intelligibility, so that a chain gain is formed.

An apparatus for generating a bandwidth extended signal from an input signal includes a patch generator and a combiner. The input signal is represented for first and second bands by first and second resolution data, respectively, the second resolution being lower than the first. The patch generator generates first and second patches from the first band of the input signal according to first and second patching algorithms, respectively. A spectral density of the second patch generated using the second patching algorithm is higher than a spectral density of a first patch generated using the first patching algorithm. The combiner combines both patches and the first band of the input signal to obtain the bandwidth extended signal. The apparatus scales the input signal according to the first and second patching algorithms or scales the first and second patches, so that the bandwidth extended signal fulfills a spectral envelope criterion.

This application provides an audio comparison method, apparatus, device, computer-readable storage medium, and product. The method includes: acquiring a first audio channel and a second audio channel; sampling the first and second audio channels respectively to obtain a first frame signal set and a second frame signal set; determining the spectral envelope window of each frame signal in the second frame signal set based on the second frame signal set and a preset critical bandwidth; shifting the spectral envelope window of each frame signal in the second frame signal set to the spectral envelope window of the corresponding frame signal in the first frame signal set to obtain the audio similarity between each frame signal in the second frame signal set and the corresponding frame signal in the first frame signal set; and determining the audio comparison result of the first and second audio channels based on the audio similarity. This application's embodiments transform the tedious audio testing task into a fully automated comparison test, improving testing efficiency and accuracy.

An audio processing method includes: filtering an audio signal to obtain a low-frequency signal and a high-frequency signal; encoding the low-frequency signal to obtain a bitstream of the low-frequency signal; performing frequency domain transform on the low-frequency signal and the high-frequency signal respectively, to obtain a low-frequency spectrum and a high-frequency spectrum; performing spectral envelope extraction on the low-frequency spectrum and the high-frequency spectrum to obtain spectral envelope information, and performing spectral flatness extraction on the high-frequency spectrum to obtain spectral flatness information; and performing quantization encoding on the spectral flatness information and the spectral envelope information to obtain a bandwidth extension bitstream, and combining the bandwidth extension bitstream and the bitstream of low-frequency signal into an encoded bitstream.

An echo suppressing device includes: an echo canceller that suppresses a linear echo signal from an input signal acquired by a microphone; a spectrum envelope extraction unit that extracts spectrum envelope information from a reception signal to be output to a loudspeaker; a nonlinear echo estimation unit that estimates spectrum envelope information of a nonlinear echo signal included in the input signal from the spectrum envelope information extracted from the reception signal; and a nonlinear echo suppression unit that suppresses the nonlinear echo signal from an output signal of the echo canceller by using the estimated spectrum envelope information of the nonlinear echo signal.

The present document relates an audio encoding and decoding system (referred to as an audio codec system). In some embodiments, a method of audio signal encoding comprises: receiving an input audio signal; transforming a sequence of samples of the input audio signal into a block of transform coefficients, the transform coefficients indicative of the spectral energy of the block; estimating a spectral envelope of the block from the transform coefficients; adjusting the transform coefficients using the spectral envelope and a spectral shaper, the spectral shaper including one or more parameters indicative of a fundamental frequency of a multi-sinusoidal signal model, where the fundamental frequency corresponds to a time domain delay; and entropy coding the adjusted transform coefficients.

This invention discloses a flat multi-wavelength vector vortex fiber laser, which utilizes a dual-injection recirculating frequency-shifting fiber cavity and a few-mode long-period fiber grating to achieve the generation of a flat multi-wavelength vector vortex beam. The laser mainly includes an injection seed source, a pump source, a wavelength division multiplexer, an erbium-doped fiber, an RF drive signal, a phase modulator, a few-mode long-period fiber grating, a partial mirror, a fiber delay unit, a fiber circulator, a polarization controller, and a fiber coupler. This laser uses a recirculating frequency-shifting fiber cavity to suppress mode competition in the erbium-doped fiber, achieving multi-wavelength oscillation. By injecting two seed lights externally, it generates complementary triangular spectral envelopes, achieving a flat multi-wavelength vector vortex beam output between the injected wavelengths. Since this laser does not use an interferometric filter as a multi-wavelength oscillation device, it exhibits excellent time stability.

The invention discloses an electromagnetic compatibility intelligent detection system and method based on an intelligent sensor, and relates to the technical field of electromagnetic compatibility detection, and the system comprises an input source collection module, a prediction analysis module and an interference recognition module. The method is technically characterized by comprising the following steps: acquiring the environment temperature of a to-be-detected area, and synchronously acquiring historical calibration data of an active antenna; wherein the historical calibration data at least comprises an electromagnetic near-field distribution map and a spectrum envelope feature; based on the historical calibration data, establishing a logic regression model to predict a drift rule of the active antenna along with the environment temperature, and obtaining a gain change curve; a one-dimensional convolution analysis electromagnetic signal is introduced to identify an electromagnetic interference event, a built-in rule engine is used for executing judgment analysis on measurement gain deviation of each batch, and an adjustment strategy is triggered through the electromagnetic interference event; according to the invention, intelligent detection of electromagnetic compatibility is realized, real-time calibration and correction are carried out in the detection process, and the detection precision is improved.

This invention relates to the field of audio processing technology, and more particularly to an audio optimization method, apparatus, computer device, and medium based on artificial intelligence. The method uses a linear layer to map the spectral envelope of the audio to be optimized to obtain envelope features; uses an embedding layer to embed standard audio parameters as parametric features; uses a prediction model to predict the fusion features of the envelope features and parametric features to obtain a predicted pitch curve; uses a noise-adding model to add noise to the Mel spectrum of the audio to be optimized to obtain a noise-adding result; uses a noise estimation model to calculate the noise of the noise-adding result to obtain predicted noise; updates the noise estimation model based on the predicted noise, the actual noise, and the predicted pitch curve; uses the updated noise estimation model to calculate the reference noise of the noise-adding result; and denoises the noise-adding result based on the reference noise to obtain the optimized Mel spectrum. Combining pitch information with the optimized noise estimation model ensures that the denoising process meets pitch requirements, thus improving the audio optimization effect.

The present invention discloses a real-time online detection and real-time feedback system for speech clarity, comprising a microphone for real-time acquisition of speech signals; an embedded computer for real-time short-time Fourier transform of an active speech signal extracted from the speech signal, dividing the obtained active speech spectrum into key frequency bands and calculating the spectrum envelope and average peak spectrum of each key frequency band, performing low-pass filtering on the average peak spectrum and then performing multi-resolution discrete Fourier transform, dividing the transformed result into fluctuation frequency bands and calculating the fluctuation spectrum mean of each fluctuation frequency band; calculating the speech modulation degree of the fluctuation frequency band according to the fluctuation spectrum mean, performing auditory correction on the speech modulation degree and then calculating the signal-to-noise ratio; calculating the transfer index, the modulation transfer index and the speech transfer index in sequence according to the signal-to-noise ratios of all fluctuation frequency bands; and displaying the speech transfer index as a speech clarity analysis result in a graphical manner.

The invention provides a foreign language learning loud-reading quality analysis system and method based on voice interaction, and the method comprises the steps: collecting a loud-reading audio signal of a learner, extracting a fundamental frequency contour and a spectrum envelope feature through a Mel-frequency cepstral coefficient algorithm, and obtaining a preliminary voice feature set; aligning duration information and energy distribution of the learner audio and the standard reference audio by adopting a dynamic time warping algorithm according to the preliminary voice feature set, and determining an aligned deviation vector; processing the combination of rhythm adjustment indexes and the preliminary voice feature set by adopting a long-short-term memory network to obtain a standardized feature vector representation form; aiming at the standardized feature vector representation form, through calculating the Euclidean distance between the standardized feature vector representation form and a standard pronunciation vector, determining a cross-language comparison difference degree; and generating an adaptive improved sequence according to the targeted feedback data to obtain a final reading quality analysis result. According to the method, through fusion of rhythm adjustment and cross-language difference analysis, the accuracy and personalized improvement effect of reading quality evaluation are remarkably improved, and efficient technical support is provided for language learning.

The invention relates to the technical field of voice processing, and particularly discloses a PTE spoken language question and answer evaluation method and system based on intelligent voice, and the method comprises the steps: obtaining an original voice signal, and extracting an acoustic feature sequence; calculating an instantaneous speech speed through phase-space reconstruction and a recursion rate matrix, and generating a time length regularity factor when the instantaneous speech speed exceeds a threshold value; determining a fractional Fourier transform order according to the factor, calculating a cosine distance matrix weighted by the instantaneous frequency deviation, and dynamically planning and backtracking to obtain a regular feature sequence; performing multi-fractal detrending fluctuation analysis to position a singular fluctuation candidate section, judging a fuzzy phoneme section by combining a certainty percentage and a laminar flow percentage of a recurrence plot, and extracting a phoneme label of an adjacent stable section as a reconstruction condition; performing phase reconstruction on the clear feature sequence, calculating a distance between a phase cross-correlation peak value and a spectral envelope Riemann geodesic line to obtain a pronunciation confidence coefficient, and performing weighted summation on the pronunciation confidence coefficient and a fluency parameter to obtain an evaluation score; according to the method, the problem of score distortion caused by duration mismatch and continuous reading boundary fuzziness at the supernormal speech speed is solved.

An approximation-free and iteration-free method for spectral analysis of an intracavity electro-optic modulation type optical frequency comb, includes: calculating a residual phase delay of a single propagation of laser in a resonant cavity, analyzing outgoing transmission characteristics of a light source of the intracavity electro-optic modulation type optical frequency comb, accumulating laser electric field intensities corresponding to all cyclic propagation times n to obtain an outgoing laser electric field intensity E, obtaining a new approximate-free outgoing laser electric field intensity E′ of the intracavity electro-optic modulation type optical frequency comb, obtaining an outgoing laser electric field intensity Ek′ of kth-order comb teeth, calculating an outgoing laser light intensity Ik of the kth-order comb teeth and accurately analyzing a spectrum of the intracavity electro-optic modulation type optical frequency comb, determining a working state according to a simulated spectral envelope curve, and guiding the subsequent optimization design and debugging.

The invention discloses a regional event detection method based on a distributed acoustic sensing system, and belongs to the technical field of event detection. According to the invention, a self-adaptive multi-scale threshold is adopted in the process of processing an initial event detection fusion signal to obtain a final event fusion signal, and a wavelet coefficient is corrected by using the self-adaptive multi-scale threshold; the corrected wavelet coefficient is used for reconstructing the signal to obtain a final event fusion signal, so that residual noise can be further eliminated, fine signal features possibly lost due to threshold processing can be recovered, and the overall quality and robustness of signal reconstruction are remarkably improved. According to the method, tuning factors are added in the feature extraction process, and the dynamic characteristics of signals in time and frequency can be deeply mined; and meanwhile, a third-order discrete cosine transform coefficient is introduced as a compensation item for feature extraction, so that more complex and finer fluctuating changes in a frequency spectrum envelope are fitted more accurately, and feature information with higher distinction degree is provided for a subsequent classification task.

Systems and techniques are provided for processing audio data. For example, a process can include determining spectral envelope features of the audio based on processing the audio using a filter bank including a plurality of filters, and determining one or more pitch features of the audio. A first neural network-based autoencoder can be used to generate a first encoded representation corresponding to the spectral envelope features, wherein the first encoded representation is based on state feedback between a decoder portion and an encoder portion of the first neural network-based autoencoder. A second encoded representation can be generated corresponding to the one or more pitch features. The first encoded representation corresponding to the spectral envelope features and the second encoded representation corresponding to the one or more pitch features can be transmitted to a decoder.

The invention discloses an optical frequency comb generating device based on a thin-film lithium tantalate grating FP (Fabry-Perot) microcavity, which is characterized by comprising a pumping unit, a microwave modulation unit and a normal dispersion grating FP resonant cavity unit. A low-power / high-power single-frequency laser pump enters an input coupling film lithium tantalate waveguide, and then is coupled to a normal dispersion grating FP resonant cavity unit through an Euler bending-based S-shaped waveguide and a small-perimeter annular micro-cavity, effective excitation of a fundamental mode is ensured based on Euler bending, a high-order mode is suppressed, a sideband is generated through electro-optical modulation, and a high-frequency laser is generated. Multi-frequency laser is generated, spectrum broadening and spectrum envelope planarization of the optical frequency comb are achieved through the combined effect of normal dispersion, the electro-optical modulation effect, the third-order nonlinear effect and high-reflectivity broadband apodized grating filtering, and finally the generated optical frequency comb is output through the output coupling waveguide. Complex mode cross coupling regulation and control are avoided, and generation of a high-efficiency optical frequency comb with a relatively flat normal dispersion area is realized.

Systems and techniques are provided for processing audio data. For example, a process can include determining pitch information associated with audio and determining spectral envelope features of the audio, wherein the spectral envelope features do not include a representation of the pitch information of the audio. The process can include determining one or more energy features of the audio. The process can include generating a combined feature vector associated with the audio and indicative of the pitch information, the spectral envelope features, and the one or more energy features. A neural network-based autoencoder can generate an encoded representation of the combined feature vector, based on state feedback between a decoder portion and an encoder portion of the neural network-based autoencoder. The encoded representation of the combined feature vector associated with the audio can be transmitted to a decoder.

The embodiment of the invention provides a frequency spectrum control method, circuit, chip and system for an ultrasonic signal. The method comprises the following steps: generating a frequency spectrum envelope corresponding to a frequency spectrum envelope parameter according to the frequency spectrum envelope parameter; according to a preset frequency point selection strategy, sequentially selecting a plurality of resident frequency points between a starting frequency point and a cut-off frequency point of the spectrum envelope to obtain a resident frequency point sequence; and determining the residence period of each residence frequency point in the residence frequency point sequence according to the total number of the residence periods and the spectrum envelope, and obtaining a frequency hopping pattern composed of the residence frequency point sequence and the residence periods. In the embodiment of the invention, only a small amount of frequency spectrum envelope parameters need to be stored, the recovery of the frequency spectrum envelope is realized through the frequency spectrum envelope parameters during use, and the signal with the corresponding frequency spectrum envelope shape can be generated based on the corresponding resident frequency point and resident period selection strategy. Therefore, storage resources are saved, flexible spectrum control can be realized, and the performance and the application range of the ultrasonic detection system are further improved.