Showing papers on "Spectrogram published in 1999"

Recent progress toward real-time measurement of ultrashort laser pulses

Abstract: Frequency-resolved optical gating (FROG) is a technique that produces a spectrogram of an ultrashort laser pulse optically. While a great deal of information about the pulse can be gleaned from its FROG trace, often it is desirable to obtain of the pulse information immediately, in real time. Quantitative information about the guise is not readily obtainable from its spectrogram without the use of a two-dimensional phase retrieval algorithm. While current algorithms are quite robust, retrieval of all the pulse information can be slow. In this paper, I describe a recently developed FROG trace inversion algorithm called Principal Component Generalized Projects that is fast, robust, and can invert FROG traces in real time. A femtosecond oscilloscope based on second-harmonic generation FROG is also described that uses this new algorithm to rapidly (up to 2.3 Hz) and continuously display the intensity and phase of ultrashort laser pulses.

A dynamic system approach to speech enhancement using the H/sub /spl infin// filtering algorithm

Abstract: This paper presents a new approach to speech enhancement based on the H/sub /spl infin// filtering. This approach differs from the traditional modified Wiener/Kalman filtering approach in the following two aspects: (1) no a priori knowledge of the noise source statistics is required, the only assumption made is that noise signals have a finite energy; (2) the estimation criterion for the filter design is to minimize the worst possible amplification of the estimation error signals in terms of the modeling errors and additive noise. Since most additive noise in speech are nonGaussian, this estimation approach is highly robust and more appropriate in practical speech enhancement. The proposed approach is straightforward to implement, as detailed in this paper. Experimental results show consistently superior enhancement performance of the H/sub /spl infin// filtering algorithm over the Kalman filtering counterpart, measured by the global signal-to-noise ratio (SNR). Examination of the spectrogram displays for the enhanced speech shows that the H/sub /spl infin// filtering approach tends to be more effective where the assumptions on the noise statistics are less valid.

Kalman-filtering speech enhancement method based on a voiced-unvoiced speech model

Abstract: In this work, we are concerned with optimal estimation of clean speech from its noisy version based on a speech model we propose. We first propose a (single) speech model which satisfactorily describes voiced and unvoiced speech and silence (i.e., pauses between speech utterances), and also allows for exploitation of the long term characteristics of noise. We then reformulate the model equations so as to facilitate subsequent application of the well-established Kalman filter for computing the optimal estimate of the clean speech in the minimum-mean-square-error sense. Since the standard algorithm for Kalman-filtering involves multiplications of very large matrices and thus demands high computational cost, we devise a mathematically equivalent algorithm which is computationally much more efficient, by exploiting the sparsity of the matrices concerned. Next, we present the methods we use for estimating the model parameters and give a complete description of the enhancement process. Performance assessment based on spectrogram plots, objective measures and informal subjective listening tests all indicate that our method gives consistently good results. As far as signal-to-noise ratio is concerned, the improvements over existing methods can be as high as 4 dB.

Multiple window time-frequency distribution and coherence of EEG using Slepian sequences and Hermite functions

Abstract: Multiple window (MW) time-frequency analysis (TFA) is a newly developed technique to estimate a time-varying spectrum for random nonstationary signals with low bias and variance. In this paper, the authors describe the application of MW-TFA techniques to electroencephalogram (EEG) and compare the results with those of the conventional spectrogram. They find that the MW-TFA provides them with not only low bias and variance time-frequency (TF) distribution for EEG but also TF coherence estimation between a single realization of EEG recorded from two sites. They also compare the performance of the MW-TFA using two sets of windows, Slepian sequences, and Hermite functions. If care is taken in matching the two windows, the authors find no noticeable difference in the resulting TF representations.

System for sound file recording, analysis, and archiving via the internet for language training and other applications

Abstract: The invention is a system for sound file recording, comparison, and archiving for network-based language and communications training, or other applications. The invention allows capture of multimedia data from a user, and allows the user to play back his or her self-created sound inputs and to view various comparisons of his or her sound inputs with model sounds. The invention displays a waveform or spectrogram of a model sound superimposed over a waveform (or spectrogram) of the user's sound input. It can display a failure/success indication for the user's sound input relative to a predetermined standard. Further, the invention allows a user to archive sound files for subsequent review and analysis.

Sound source classification

Abstract: A system and method to identify a sound source among a group of sound sources. The invention matches the acoustic input to a number of signal models, one per source class, and produces a goodness-of-match number for each signal model. The sound source is declared to be of the same class as that of the signal model with the best goodness-of-match if that score is sufficiently high. The data are recorded with a microphone, digitized and transformed into the frequency domain. A signal detector is applied to the transient. A harmonic detection method can be used to determine if the sound source has harmonic characteristics. If at least some part of a transient contains signal of interest, the spectrum of the signal after resealing is compared to a set of signal models, and the input signal's parameters are fitted to the data. The average distortion is calculated to compare patterns with those of sources that used in training the signal models. Before classification can occur, a source model is trained with signal data. Each signal model is built by creating templates from input signal spectrograms when they are significantly different from existing templates. If an existing template is found that resembles the input pattern, the template is averaged with the pattern in such a way that the resulting template is the average of all the spectra that matched that template in the past.

Method and apparatus for analyzing hybridized biochip patterns using resonance interactions employing quantum expressor functions

Abstract: A technique is described for identifying mutations, if any, present in a biological sample, from a pre-selected set of known mutations. The method can be applied to DNA, RNA and peptide nucleic acid (PNA) microarrays. The method analyzes a dot spectrogram representative of quantized hybridization activity of oligonucleotides in the sample to identify the mutations. In accordance with the method, a resonance pattern is generated which is representative of nonlinear resonances between a stimulus pattern associated with the set of known mutations and the dot spectrogram. The resonance pattern is interpreted to a yield a set of confirmed mutations by comparing resonances found therein with predetermined resonances expected for the selected set of mutations. In a particular example, the resonance pattern is generated by iteratively processing the dot spectrogram by performing a convergent reverberation to yield a resonance pattern representative of resonances between a predetermined set of selected Quantum Expressor Functions and the dot spectrogram until a predetermined degree of convergence is achieved between the resonances found in the resonance pattern and resonances expected for the set of mutations. The resonance pattern is analyzed to a yield a set of confirmed mutations by mapping the confirmed mutations to known diseases associated with the pre-selected set of known mutations to identify diseases, if any, indicated by the biological sample. By exploiting a resonant interaction, mutation signatures may be robustly identified even in circumstances involving low signal to noise ratios or, in some cases, negative signal to noise ratios.

Method and apparatus for interpreting hybridized bioelectronic DNA microarray patterns using self-scaling convergent reverberant dynamics

Abstract: A technique is described for identifying mutations, if any, present in a biological sample, from a pre-selected set of known mutations. The method can be applied to DNA, RNA and peptide nucleic acid (PNA) microarrays. The method analyzes a dot spectrogram representative of quantized hybridization activity of oligonucleotides in the sample to identify the mutations. In accordance with the method, a resonance pattern is generated which is representative of nonlinear resonances between a stimulus pattern associated with the set of known mutations and the dot spectrogram. The resonance pattern is interpreted to a yield a set of confirmed mutations by comparing resonances found therein with predetermined resonances expected for the selected set of mutations. In a particular example, the resonance pattern is generated by iteratively processing the dot spectrogram by performing a convergent reverberation to yield a resonance pattern representative of resonances between a predetermined set of selected Quantum Expressor Functions and the dot spectrogram until a predetermined degree of convergence is achieved between the resonances found in the resonance pattern and resonances expected for the set of mutations. The resonance pattern is analyzed to a yield a set of confirmed mutations by mapping the confirmed mutations to known diseases associated with the pre-selected set of known mutations to identify diseases, if any, indicated by the biological sample. By exploiting a resonant interaction, mutation signatures may be robustly identified even in circumstances involving low signal to noise ratios or, in some cases, negative signal to noise ratios.

Spectrogram enhancement algorithm: a soft thresholding-based approach

Abstract: Enhancing the spectrogram by denoising the Doppler ultrasound signal is a preliminary step, and important for further processing. Because the spectrogram may be based on the short-time fast Fourier transform (FFT) of the Doppler ultrasound signal, whose power spectrum density is time-varying, traditional denoising algorithms that simply optimize the mean-squared error are not appropriate, and they may exhibit considerable undesirable, noise-induced frequency components. A soft thresholding-based denoising algorithm is put forward in this paper, that achieves almost the minimax mean square error (MSE) over a wide range of function classes having norms measuring smoothness (i.e., it meets both the requirement of smoothness and MSE). Due to the importance of noise level estimation while applying this method, several robust L-estimators are compared and the median absolute deviation (MAD) method is chosen to estimate the noise level. The simulation study shows better performance of the later algorithm under various quantification measures, compared to the FFT thresholding and the hard thresholding wavelet method, and the results of clinical data also confirm it.

The use of a filter bank and the Wigner-Ville distribution for time-frequency representation

Abstract: We present a new method for time-frequency representation, which combines a filter bank and the Wigner-Ville distribution (WVD). The filter bank decomposes a multicomponent signal into a number of single component signals before the WVD is applied. Cross-terms as well as noise are reduced significantly, whereas high time-frequency concentration is attained. Properties of the proposed time-frequency distribution (TFD) are investigated, and the requirements for the filter bank to fulfil these are given. The ability of the proposed non-Cohen's (1995) class TFD to reduce cross-terms as well as noise as well as its ability to approximately reconstruct signals are illustrated by examples. The results are compared with those from the WVD, the Choi-Williams (1989) distribution (CWD), and spectrogram.

Spectrographic measurement of instantaneous frequency and the time-dependent weighted average instantaneous frequency

Abstract: Instantaneous frequency (IF) is an important signal characteristic arising in many fields. It is a concept intimately linked to time–frequency analysis, where it can be obtained from a time–frequency distribution (TFD) as the first conditional moment in frequency, suggesting that the IF is the average frequency at each time. However, this interpretation is questionable, since it is well known that the IF often ranges beyond the spectral support of the signal. In addition, to obtain the IF from a spectrogram (which is one possible TFD), a very wideband—and thus severely spectrally distorted—spectrogram must be used. More reasonable bandwidths are investigated, and, in particular, give the conditions by which the first conditional moment of a spectrogram can indeed be interpreted as the average frequency at each time. Under these conditions, namely when the spectrogram bandwidth is less than the frequency separation of the signal components in the time–frequency plane, the spectrogram yields not the usual IF, but a time-dependent weighted average instantaneous frequency (WAIF) of the signal. Also, while the IF and WAIF are generally different, sometimes they are the same (in particular, when there is symmetry in the time–frequency spectrum of the signal); in that case, the first conditional spectral moments of both wideband and narrow band spectrograms are the same and interpretable as the average frequency at each time.

Automatic evaluation of plasma density profiles from microwave reflectometry on ASDEX upgrade based on the time-frequency distribution of the reflected signals

Abstract: A new data analysis technique has been developed for the evaluation of the density profiles from broadband reflectometry in the presence of plasma turbulence. The method is based on the spectrogram of the reflected signals and uses the complete information of the beat frequency spectrum. The application of the Floyd best path algorithm, that takes into account the history of the beat frequency curve, enables us to extract the slow component due to the plasma profile. The statistics of the group delay data points is also considered to validate (or reject) data with great confidence. The results obtained in a wide range of plasma regimes show that accurate and detailed profiles can be measured automatically. One and two dimensional moving averaging over consecutive sweeps is also available to reduce the variance of the inverted profile while retaining a good temporal resolution.

The Wigner distribution as a tool for time-frequency analysis of fusion plasma signals: application to broadband reflectometry data

Abstract: The Wigner distribution is introduced as an appropriate tool for processing data from fusion plasma diagnostics whose signals have a time varying frequency spectrum, and is thus presented as a particularly suited form for the time-frequency analysis of transients and other non-stationary phenomena. Its effectiveness is illustrated by applying it to the problem of electron density profile measurement through broadband microwave reflectometry in fusion devices, its advantages being demonstrated via a detailed comparison with the well known short time Fourier transform spectrogram. In particular, the Wigner distribution is used in a novel application as a means not only to retrieve from reflectometry data the instantaneous frequency needed for profile inversion, but also to provide an accurate representation of reflectometry signals in the time-frequency plane. Further, in a careful discussion stressing its benefits relative to more standard approaches based on the quadrature signal, the analytic signal is proposed as being appropriate to routinely extract from reflectometry data an unambiguously defined set of instantaneous amplitude, phase and frequency. The problems associated with the fact that digitized reflectometry data give rise to discrete time signals are properly handled, and the different estimates for the instantaneous frequency obtained from the Wigner distribution and the spectrogram by calculating their first frequency moments and mean frequencies, as well as by locating their peaks, are also compared.

Spectrogram analysis of selected tremor signals using short-time Fourier transform and continuous wavelet transform

Abstract: Among a variety of spectrogram methods Short-Time Fourier Transform (STFT) and Continuous Wavelet Transform (CWT) were selected to analyse transients in non-stationary tremor signals. Depending on the properties of the tremor signal a more suitable representation of the signal is gained by CWT. Three selected broadband tremor signals from the volcanos Mt. Stromboli, Mt. Semeru and Mt. Pinatubo were analyzed using both methods. The CWT can also be used to extend the definition of coherency into a time-varying coherency spectrogram. An example is given using array data from the volcano Mt. Stromboli.

Optical Spectrogram Scope Using Time-to-Two-Dimensional Space Conversion and Interferometric Time-of-Flight Cross Correlation

Abstract: We demonstrate a new tool called optical spectrogram scope for visualization of a spectrogram or a scalogram of optical ultrafast phenomena. The optical spectrogram scope is constructed on the basis of the time-to-two-dimensional-space conversion technique capable of converting a set of time-varying frequency distributions into two-dimensional spatial ones.

Lag-windowing and multiple-data-windowing are roughly equivalent for smooth spectrum estimation

Abstract: There is no fundamental difference between lag-windowing a correlation sequence and multiple-windowing a data sequence when the objective is to reduce the mean-squared error of a spectrum estimator. By analyzing the approximate low-rank factorization of a bandlimiting Toeplitz operator, we find that lag-windowed (or spectrally smoothed) spectrum estimators have multiple-data-windowed implementations. This makes the Blackman-Tukey-Grenander-Rosenblatt spectrogram equivalent to the Thomson spectrum estimator (and vice-versa), meaning BTGR spectrograms may be implemented in a multichannel filterbank version of the Thomson estimator.

Data-driven modulation filter design under adverse acoustic conditions and using phonetic and syllabic units.

Abstract: Constructing speech feature extraction methods that are robust to many types of corrupting acoustic environments remains a daunting task and it is instructive to investigate which properties of the speech carry the discriminative information for recognition under a variety of conditions. In this paper we describe results for generating RASTA-style modulation filters under a number of acoustic environments. We utilize Linear Discriminant Analysis in a manner previously described by van Vuuren and Hermansky to automatically generate discriminant filters for speech with artificially added background noise and reverberation. We also generate the filters using both phonetic and syllabic classification targets. Trends in the responses of the discriminant filters lend support to feature extraction design decisions employed by RASTA-PLP and Modulation-filtered Spectrogram features. Further, tests with added reverberation corroborate views on the perceptual stability of syllabic rates.

Nonstationary signal analysis of magnetic islands in plasmas

Abstract: Rotating magnetic islands produce fluctuations on a variety of diagnostics in magnetic fusion energy plasmas. The analysis of these fluctuations requires the calculation of the amplitude, phase, and frequency of the oscillations. These three spectral quantities generally evolve in time, necessitating nonstationary signal analysis techniques. The Hilbert transform offers an efficient and accurate method of calculating these three quantities from one diagnostic signal. This feature allows the Hilbert transform to determine the success of the active rotation control of magnetic islands, and to calculate the profile of the diagnostic measurements in a frame of reference co-rotating with the magnetic island. Comparisons to quadrature and spectrogram techniques demonstrate the accuracy of the Hilbert transform method.

Reduction of Musical Noise in Spectral Subtraction Method Using Subframe Phase Randomization

Abstract: The Subframe phase randomization method is applied to the spectral subtraction method to reduce the musical noise in nonvoicing region after speech enhancement. The musical noise in the spectral subtraction method is the result of the narrowband tonal components that appearing somewhat periodically in the spectrogram of unvoiced and silence regions. Thus each synthesis frame in nonvoicing region is divided into several subframes to broaden the narrowband spectrum, and then phases of silence and unvoiced regions are randomized to eliminate the tonal components in the spectrum while keeping the shape of the amplitude spectrum. Performance assessments based on visual inspection of spectrogram, objective measure, and informal subjective listening tests demonstrate the superiority of the proposed algorithm.

Reduction of musical noise in spectral subtraction method using subframe phase randomisation

Abstract: A subframe phase randomisation method is proposed and applied to the enhanced speech with spectral subtraction method to reduce musical noise in the nonvoicing region. The musical noise in the spectral subtraction method is largely due to narrowband tonal components appearing somewhat periodically in the spectrogram of unvoiced and silence regions. Thus, each synthesis frame in a nonvoicing region is divided into several subframes to broaden the narrowband spectrum, and then phases of silence and unvoiced regions are randomised to disrupt the tonal spectrum structure.

Spectrogram analysis of arterial Doppler signals for off-line automated hits detection.

Abstract: Recently, a time processing of arterial Doppler signals was proposed to detect automatically high-intensity transient signals (HITS). This technique provided satisfactory detection results, but was not always constantly accurate, particularly with high-resistance blood velocity profiles. A time-frequency processing, based on the spectrogram, is presented to detect the presence of emboli in the arterial Doppler signals. The method uses the narrow-band hypothesis and extracts the detection criterion from the time-frequency representation (TFR). A first database of 560 peripheral arterial Doppler HITS was created to study microemboli and to define the normal limits to be used in our method. A threshold was experimentally defined using this database, and then applied to 38 recordings from 12 patients. Using another database, 6 human expert Doppler users reported 140, 176, 155, 161, 161 and 146 HITS, corresponding to a total of 197 different observed HITS. When an event was detected by at least 6, 5, 4, 3, 2 and 1 of the observers, sensitivity of the automatic detection was 93.9, 91.7, 89.6, 88.7, 84.7 and 73.1%, respectively. The sensitivity of our automatic detection is, thus, highly associated with the number of observers in agreement. A preliminary experiment has been performed to test the method in the case of long recording duration. In 15 patients, 6 h 24 min of recordings have been analyzed. The proposed automated processing provided an overall sensibility of 91.5%. The present work shows that this detection scheme preserves good sensibility and improves the positive predictive value compared with the time-processing recently proposed.

Multiple-window method for obtaining improved spectrograms of signals

Abstract: There is disclosed a method for processing a time-varying signal to produce a high-resolution spectrogram that represents power as a function of both frequency and time. Data blocks of a time series, which represents of a sampled signal, are subjected to processing which results in a sequence of frequency-dependent functions referred to as eigencoefficients. Each eigencoefficient represents signal information projected onto a local frequency domain using a respective one of K Slepian sequences or Slepian functions. The spectrogram is derived from time- and frequency-dependent expansions formed from the eigencoefficients.

Music database creating device, creating method, and program recording medium thereof

Abstract: PROBLEM TO BE SOLVED: To add an identifier to music information in real time with a relatively simple processing. SOLUTION: A spectrogram is calculated by analyzing frequency of input music information 205. Edge intensity (edi) of the spectrogram in the time axis direction is calculated 207. The edge intensity (edi) is summed in the frequency axis direction to calculate an edge intensity sum ED 209. Data indicating the ED or an increase, a decrease or a continuation of the ED are added to a header of the music information as an identifier to store in a database. COPYRIGHT: (C)2000,JPO

Doppler spectral estimation using time-frequency distributions

Abstract: The time-frequency distribution (TFD) of Doppler blood flow signals is usually obtained using the spectrogram, which requires signal stationarity and is known to produce large estimation variance. This paper examines four alternative, nonstationary spectral estimators: a smoothed pseudo-Wigner distribution (SPWD), the Choi-Williams distribution (CWD), the Bessel distribution (BD), and the novel, adaptive constant-Q distribution (AQD) for their applicability to Doppler ultrasound. A synthetic Doppler signal, simulating the nonaxial and pulsatile flow of the common carotid artery, was used for quantitative comparisons at different signal-to-noise-ratios (SNR) of 0, 10, 20, and 30 dB as well as noise free. The cross-correlation (/spl rho/) and the root-mean-square-error (RMSE) were calculated after log-compression for each technique and SNR relative to the theoretical distribution. The AQD consistently had the lowest RMSE (/spl les/53.7%) and the highest /spl rho/ (/spl ges/0.889) of all the TFDs, irrespective of the SNR. The SPWD performed better than the spectrogram, which performed better than the BD and the CWD. Qualitative comparisons were carried out using in vivo data acquired with a 10 MHz ultrasound cuff transducer positioned around the distal aorta of a rabbit. In vivo, the AQD was considered best with respect to background noise and internal gray scale features; it was rated second (after the spectrogram) in depicting the spectral envelope. The AQD performed better as a Doppler spectral estimator than the traditional spectrogram and the other TFDs under the conditions studied here.

Signal analyzer system and method for computing a fast Gabor spectrogram

Abstract: A signal analyzer, method and memory medium for generating a time varying spectrum for input signals characterized by frequency components which change in time. The signal analyzer includes a source of a sequence of digital signals representative of an input signal, a processor coupled to the source, and a memory medium coupled to the processor. The memory medium stores a software program which is executable by the processor to compute the time-varying spectrum of the input signal. When the processor executes the software program, the processor is operable to first compute a Gabor transform (that is, a sampled short-time Fourier transform) of the digital signals to produce Gabor coefficients. The processor then computes a two dimensional auto-correlation of the Gabor coefficients to produce auto-correlation results. The auto-correlation results are then applied to a 2-dimensional fast interpolation filter to produce the time-varying spectrum, wherein the time-varying spectrum is a Gabor spectrogram. The signal analyzer may repeat the above steps n+1 times, based on the order determined by a user, and sum the results for an n order time-varying spectrum. The process more may then operate to process and/or display the time-varying spectrum.

A 4 kbps adaptive fixed code-excited linear prediction speech coder

Abstract: We propose an adaptive fixed code-excited linear prediction (AF-CELP) speech coder operating at 4 kbps. By exploiting the fact that a fixed codebook contribution to the speech signal is also periodic as the corresponding adaptive codebook contribution, the adaptive fixed codebook model efficiently represents excitation signals. In order to overcome the quality degradation caused by the coarse quantization of excitation, a paired pulse algebraic codebook structure is also applied to the excitation model. Additionally, a pitch prefiltering, a noise spreading, and a harmonic enhancement technique are adopted in the decoding process. The spectrogram reading and informal listening tests proved that the AF-CELP reproduces high quality speech.

A new multi-window time-frequency approach yielding accurate low-order conditional moments

Abstract: The authors build on Cohen's work on instantaneous bandwidth and frequency by extending it to a multi-window framework. Unlike the case with a single spectrogram, the multi-window framework allows one to obtain a time-varying spectral estimate that simultaneously satisfies instantaneous bandwidth and frequency constraints.