Adaptive WRLS-VFF for speech analysis

The electronic structures of the Floquet states of the dynamic Wannier-Stark ladder (DWSL) are examined, where the DWSL is formed by driving the biased superlattices (SLs) by the periodic pulse train (PPT) with the electric field $F(t)$---with time $t$---and the temporal period $2\ensuremath{\pi}∕\ensuremath{\omega}$. For a strong $F(t)$, interminiband interactions, namely, the ac-Zener tunneling (ac-ZT), are predominantly caused in the DWSL. Such a system is termed the interacting DWSL. In order to understand the details of the Floquet states and the modulation patterns by alteration of a couple of the PPT laser parameters, the linear absorption spectra, ${\ensuremath{\alpha}}_{\mathit{abs}}({\ensuremath{\omega}}_{p};\ensuremath{\omega})$, of optical interband transitions invoked by the monochromatic probe laser ${f}_{p}(t)$ with the frequency ${\ensuremath{\omega}}_{p}$ are calculated, where the spectra are not only linear in ${f}_{p}(t)$ but also nonlinear in $F(t)$. The exciton effect is not included for the sake of simplicity. For the PPT driving with unit-pulse shapes largely deviated from the square and saw-toothed profiles, the spectra show unexpected dent structures, differing a great deal from the corresponding ac-ZT-free spectra basically similar to those of the original SLs just showing the ascending steplike structure. To deepen the understanding of this anomaly, the spectra of ${\ensuremath{\alpha}}_{\mathit{abs}}^{0}({\ensuremath{\omega}}_{p};\ensuremath{\omega})\ensuremath{\propto}\ensuremath{\partial}{\ensuremath{\alpha}}_{\mathit{abs}}({\ensuremath{\omega}}_{p};\ensuremath{\omega})∕\ensuremath{\partial}{\ensuremath{\omega}}_{p}$ are also calculated, whereby the dent structures become spectral dips showing the negative absorption. It is found that such anomalous behavior is attributed to the ac-ZT between different minibands that accompanies emission/absorption of the nonzero net number of photons with $J\ensuremath{\omega}$ (with $J$ a nonzero integer). This anomaly also shows the unusual time dependence in the dual-time optical susceptibility associated with ${\ensuremath{\alpha}}_{\mathit{abs}}^{0}({\ensuremath{\omega}}_{p};\ensuremath{\omega})$. Moreover, the possibility of existence of the negative absorption in the more realistic excitonic spectra is speculated.

Poor sensitivity to the interaural time difference (ITD) constrains the ability of human bilateral cochlear implant users to listen in everyday noisy acoustic environments. ITD sensitivity to periodic pulse trains degrades sharply with increasing pulse rate but can be restored at high pulse rates by jittering the interpulse intervals in a binaurally coherent manner (Laback and Majdak. Binaural jitter improves interaural time-difference sensitivity of cochlear implantees at high pulse rates. Proc Natl Acad Sci USA 105: 814-817, 2008). We investigated the neural basis of the jitter effect by recording from single inferior colliculus (IC) neurons in bilaterally implanted, anesthetized cats. Neural responses to trains of biphasic pulses were measured as a function of pulse rate, jitter, and ITD. An effect of jitter on neural responses was most prominent for pulse rates above 300 pulses/s. High-rate periodic trains evoked only an onset response in most IC neurons, but introducing jitter increased ongoing firing rates in about half of these neurons. Neurons that had sustained responses to jittered high-rate pulse trains showed ITD tuning comparable with that produced by low-rate periodic pulse trains. Thus, jitter appears to improve neural ITD sensitivity by restoring sustained firing in many IC neurons. The effect of jitter on IC responses is qualitatively consistent with human psychophysics. Action potentials tended to occur reproducibly at sparse, preferred times across repeated presentations of high-rate jittered pulse trains. Spike triggered averaging of responses to jittered pulse trains revealed that firing was triggered by very short interpulse intervals. This suggests it may be possible to restore ITD sensitivity to periodic carriers by simply inserting short interpulse intervals at select times.

Sensitivity to interaural time differences (ITDs) in high-frequency bandpass-filtered periodic and aperiodic (jittered) pulse trains was tested at a nominal pulse rate of 600 pulses per second (pps). It was found that random binaurally-synchronized jitter of the pulse timing significantly increases ITD sensitivity. A second experiment studied the effects of rate and place. ITD sensitivity for jittered 1200-pps pulse trains was significantly higher than for periodic 600-pps pulse trains, and there was a relatively small effect of place. Furthermore, it could be concluded from this experiment that listeners were not solely benefiting from the longest interpulse intervals (IPIs) and the instances of reduced rate by adding jitter, because the two types of pulse trains had the same longest IPI. The effect of jitter was studied using a physiologically-based model of auditory nerve and brainstem (medial superior olive neurons). It was found that the random timing of the jittered pulses increased firing synchrony in the auditory periphery, which caused an improved rate-ITD tuning for the 600-pps pulse trains. These results suggest that a recovery from binaural adaptation induced by temporal jitter is possibly related to changes in the temporal firing pattern, not spectral changes.

The excitation of photon echo by noise pulses that are formed by modulation of the carrying frequency with Gauss noise is modeled. The modeling is based on optical Bloch equations the solution of which for noise pulse realizations is constructed by their stepwise approximation. In terms of the formalism of state transfer matrices, the two- and three-pulse excitation modes are analyzed. The complex envelopes of the primary and stimulated echo responses are determined. In the linear (low-level-signal) mode, the shape of the two-pulse echo corresponds to that of the time delayed and inverted noise pulse. The boundary of the linear mode, upon exceeding of which distortions of the shape of the noise pulse become noticeable, is determined. The shape of the stimulated (three-pulse) echo in the linear mode corresponds to that of the autocorrelation function of the noise pulse realization. Upon passage beyond the boundary of the linear mode, the shape of the three-pulse echo corresponds either to the cross-correlation function of distorted noise pulses (with different intensities) or to the autocorrelation function of distorted pulses (with the same intensities). The modeled photon echo excitation modes can be used in photon echo processors to process signals in the light range.

This paper presents an example of power analysis in parallel resonant circuit excited by current defined as periodic pulse train. The cyclostationary methodology being developed into approach to the electric power analysis, which consists in exploiting a standard set of cyclostationary characteristics such as cross correlation function and its frequency counterpart – cross spectral correlation density (CSCD). These characteristics make it possible to calculate all commonly used power quantitative measures including instantaneous, apparent and average power, power factor and others. The harmonic power distribution is investigated for the chosen low-Q resonant case. The expression for components of CSCD via current and voltage Fourier series coefficients is obtained using the known frequency response of the circuit under study.

The oscillators star-coupled by one resistor exhibit the phase-shift synchronization and there exist many different steady phase patterns according to their initial states. Therefore we have suggested that these systems can be use as structural elements of neural networks and large scale memories. In such cases, to get appropriate phase patterns, controlling the phase patterns should be an important problem. In this study, we propose the Wien-bridge oscillators star-coupled by one resistor driven by the periodic pulse train. In this system, the phase pattern switching is caused by the stimulation of the pulse. From the results, we suggest that the pulse stimulation is one of the ways to control the phase patterns in coupled oscillators systems.

It was recently shown that a periodic pulse train will reproduce itself at periodic distances upon propagation in a linear dispersive medium. By drawing an analogy between this temporal self-imaging problem and the well known spatial self-imaging phenomenon, we derive the necessary condition for temporal self-imaging and show that the class of temporally self-reproducing signals is much broader than the subset of periodic signals.

Semiconductor optical amplifier (SOA)-based optical timing extraction and clock-recovery schemes offer compact size, low operating power, and added possibility for integration into an all-optical circuitry. In this paper, we analyze a device that could measure the relative delay between counterpropagating pulses incident on an SOA. Unlike previous designs based on differential photodiodes, our optical pulse delay discriminator (OPDD) uses the voltage difference between two contacts on the SOA. We therefore eliminate the two photodiodes and two optical couplers, making integration far easier. After giving a qualitative description of the operation of the proposed design, we carry out a comprehensive analytical analysis of the operation of the device performance. At each stage, we demonstrate the accuracy of the derived approximate formulas. An analytical expression for the transient response of OPDD shows it to be exponential with a time constant set by the carrier-recovery lifetime. We show that it is possible to reliably measure the incident delay between counterpropagating, periodic pulse trains by measuring the mean value (or low-pass-filtered value) of the induced voltage difference. Our analysis shows the OPDD has excellent linearity.

Some signals are transmitted as periodic pulse trains. If more than one pulse train is transmitted over the same communication channel, a challenge is to separate them for source identification at the receiver. This is known as pulse train deinterleaving and is clearly a fundamental problem in the study of discrete-event systems. Frequently, the only relevant information at the receiver is the time of arrival (TOA) data, which is usually contaminated by jitter noise. Perhaps there are also missing or overlapping pulses. In this paper, we present an approach for deinterleaving pulse trains and estimating their periods using an extended Kalman filter (EKF). A naive application of EKF theory is not attractive because of discontinuities in the signal model. Here, a form of smoothing of the discontinuities is proposed so that the EKF approach becomes attractive. The advantage of this EKF approach is that it is less computationally expensive than most previously proposed methods, which are of order N/sup 2/, where N is the number of pulses being processed. The computation required here is of order N. The method proposed appears to give useful results for up to seven or so pulse trains, particularly when there is some a priori information on the pulse frequencies, which can be obtained using computations of order N log N.

Five experiments on the identifiability of synthetic vowels masked by wideband sounds are reported. In each experiment, identification thresholds (signal/masker ratios, in decibels) were measured for two versions of four vowels: a vibrated version, in which FO varied sinusoidally around 100 Hz; and a steady version, in which F0 was fixed at 100 Hz. The first three experiments were performed on naive subjects. Experiment 1 showed that for maskers consisting of bursts of pink noise, vibrato had no effect on thresholds. In Experiment 2, where the maskers were periodic pulse trains with an F0 randomly varied between 120 and 140 Hz from trial to trial, vibrato slightly improved thresholds when the sound pressure level of the maskers was 40 dB, but had no effect for 65-dB maskers. In Experiment 3, vibrated rather than steady pulse trains were used as maskers; when these maskers were at 40 dB, the vibrated versions of the vowels were slightly less identifiable than their steady versions; but, as in Experiment 2, vibrato had no effect when the maskers were at 65 dB. Experiment 4 showed that the unmasking effect of vibrato found in Experiment 2 disappeared in subjects trained in the identification task. Finally, Experiment 5 indicated that in trained listeners, vibrato had no influence on identification performance even when the maskers and the vowels had synchronous onsets and offsets. We conclude that vibrating a vowel masked by a wideband sound can affect its identification threshold, but only for tonal maskers and in untrained listeners. This effect of vibrato should probably be considered as a Gestalt phenomenon originating from central auditory mechanisms.

Perception of temporal acoustic patterns by the goldfish ( Carassius auratus)

Detecting randomly modulated pulses in noise

ABSTRACTThis paper describes a method of extracting time-varyingfeatures that is effective for speech signals with high funda-mental frequencies. The proposed method adopts a speechproduction model that consists of a Time-Varying Auto-Regressive (TVAR) process for an articulatory filter and aHidden Markov Model (HMM) for an excitation source.The model represents waveform amplitude variations bytime-varying gain of the excitation source. The proposed algo-rithm is given by extending a Viterbi algorithm so that theproposed algorithm can adaptively estimate TVAR coeffi-cients and time-varying gain with decoding the state tran-sition of the excitation source HMM. We applied the pro-posed method to extracting time-varying features from bothsynthetic and natural speech, and confirmed its feasibility.1. INTRODUCTIONThe conventional Linear Prediction (LP) method is widelyused to analyze speech signals[1]. However, several prob-lems still remain to be solved[2]. One such problem is thatlocal peaks of the LP spectral estimate are strongly biasedtoward the harmonics, especially for high-pitched speech.Several methods have been designed to overcome this prob-lem [3, 4, 5, 6]. The authors have previously indicated thatan analysis method based on a speech production modelconsisting of an Auto-Regressive (AR) process for an artic-ulatoryfilterand a HiddenMarkov Model(HMM)for an ex-citationsource is robustforhighfundamental frequencies[7,8]. However, this method is not suitable for analyzing con-tinuous speech for the following reasons. First, the AR co-efficients and HMM are iteratively estimated within everyanalysis frame, so a large number of operations is needed.Second, the analysis frame size needs to be large in orderto guarantee stable learning of the excitation source HMM.Third, the model parameters are assumed to be constantwithin the analysis frame, so the resulting parameters areaveraged within such a long analysis frame. This makesit difficult to extract the dynamic characteristics of speechwhen features change rapidly, like in a singing voice.In this paper, we extend the speech production model in[7] so that the proposed model can represent time-varyingfeatures of continuous speech. We also describe an anal-ysis method that adaptively estimates Time-Varying Auto-Regressive (TVAR) coefficients and gain based on the newmodel. The proposed method can substantially reduce thenumber of operations by applying the learned HMM andcan also extract dynamic characteristics of continuous speechby estimating those time-varying features adaptively.2. SPEECH PRODUCTION MODEL BASED ONTVAR-HMMThe proposed method adopts a speech production modelthat consists of a TVAR process for an articulatory filterand an HMM for an excitation source. The nodes of theHMM are concatenated in a ring state in order to representperiodicity of voiced sounds. We have previously shownthat LP analysis incorporating the excitation source HMMcan precisely estimate the characteristics of both vocal tractand excitation source from high-pitched speech signal[7, 8].The proposed model represents the time-varying features ofnot only the vocal tract but also the waveform amplitude bymultiplying an excitation source emitted from the HMM bya time-varying gain.

This research aimed to investigate the accuracy of picking of P-wave arrival times in rock fracture acoustic emission signals. In order to simulate the mining scenario, Gaussian white noise and pulse noise were added to the data collected in the laboratory. Complete ensemble empirical mode decomposition with adaptive noise + Wavelet (CEEMDAN + Wavelet) was improved in this paper, where the Spearman rank correlation coefficient was adopted to effectively select intrinsic mode functions for denoising which retained the inherent characteristics of the rock fracture signal. The absolute amplitude and energy change rate of the envelope signal, calculated based on the Hilbert transform, were used as the input of the short term average/long term average (STA/LTA) normalization algorithm to pickup the P-wave arrival time. The reliability of this method was tested on 30 groups of recorded rock fracture laboratory data and 60 groups of added noise data. Taking the manual pickup results as the standard, the errors of CEEMDAN + Wavelet + STA/LTA + AIC (Akaike information criterion) method with the absolute amplitude of the signal as the input are all within 10 ms, and 86.67% of the results are within 5 ms. The method proposed in this paper effectively addressing the issue of false pickup caused by the sensitivity of AIC and traditional STA/LTA method for strong noise, and achieving relatively high accuracy and stability in processing low signal-to-noise ratio signals. This work contributes to monitor microscopic changes in rock bodies and is of great significance for the prediction and monitoring of geological disasters.

The critical phenomena in a one-dimensional periodic stratified medium with time-varying dielectric permittivity functions are theoretically investigated for both TM and TE waves by solving the time-dependent electric displacement equation and the magnetic field equation, respectively. It is found that there exists a critical value of the varying amplitude when the dielectric permittivity function is changing with time. When the varying amplitude is below or at the critical value, the periodic structure can reflect an incident plane wave into a periodic femtosecond pulse train with respect to time. On the other hand, when the varying amplitude is much above the critical value, the reflectivity is descending more rapidly with time, which means that, at one moment, the incident light is transmitted thoroughly and the periodic structure is becoming transparent. The effect of the layer number on the reflectivity is also presented. Especially at the critical value, the pulse train generated is periodic with almost 100% peaks and zero troughs for both TM and TE waves when the layer number is up to 11. These interesting conclusions could find potential realistic applications in the future.