Noisy Waveforms Research Articles

Comparison of neural network architectures for feature extraction from binary black hole merger waveforms

We evaluate several neural-network architectures, both convolutional and recurrent, for gravitational-wave time-series feature extraction by performing point parameter estimation on noisy waveforms from binary-black-hole mergers. We build datasets of 100 000 elements for each of four different waveform models (or approximants) in order to test how approximant choice affects feature extraction. Our choices include SEOBNRv4P and IMRPhenomPv3, which contain only the dominant quadrupole emission mode, alongside IMRPhenomPv3HM and NRHybSur3dq8, which also account for high-order modes. Each dataset element is injected into detector noise corresponding to the third observing run of the LIGO-Virgo-KAGRA (LVK) collaboration. We identify the temporal convolutional network architecture as the overall best performer in terms of training and validation losses and absence of overfitting to data. Comparison of results between datasets shows that the choice of waveform approximant for the creation of a dataset conditions the feature extraction ability of a trained network. Hence, care should be taken when building a dataset for the training of neural networks, as certain approximants may result in better network convergence of evaluation metrics. However, this performance does not necessarily translate to data which is more faithful to numerical relativity simulations. We also apply this network on actual signals from LVK runs, finding that its feature-extracting performance can be effective on real data.

An Experimental Electronic Board ADF339 for Analog and FPGA-Based Digital Filtration of Measurement Signals

This work introduces and examines a new programmable electronic system, Board ADF339, designed for filtering analog measurement signals of low frequencies. The system operates in a mixed mode in collaboration with a digital controller implemented on the myRIO-1900 FPGA module. It enables the digital selection of the type and frequency settings of the UAF42 integrated circuit. In the technical implementation section, electronic filter and phase shifter circuit diagrams are presented, along with the digital counterpart of the analog filter. Tests of this system were conducted on signals generated using a function generator, which was followed by the filtration of signals occurring in real laboratory setups. A series of real responses from three different laboratory systems and a measurement system utilizing LabVIEW FPGA virtual instruments are demonstrated. After computing SNR indicators for noisy waveforms, the application scope and usability of the board are highlighted.

A Joint Speech Enhancement and Self-Supervised Representation Learning Framework for Noise-Robust Speech Recognition

Though speech enhancement (SE) can be used to improve speech quality in noisy environments, it may also cause distortions that degrade the performance of automatic speech recognition (ASR) models. Self-supervised pre-training, on the other hand, has been shown to improve the noise robustness of ASR models. However, the potential of the (optimal) integration of SE and self-supervised pre-training still remains unclear. In this paper, we propose a novel self-supervised pre-training framework that incorporates SE to improve ASR performance in noisy environments. First, in the pre-training phase the original noisy waveform or the waveform obtained by SE is fed into the self-supervised model to learn the contextual representation, where the quantized clean speech acts as the target. Second, we propose a dual-attention fusion method to fuse the features of noisy and enhanced speech, which can compensate for the information loss caused by separately using individual modules. Due to the flexible exploitation of clean/noisy/enhanced branches, the proposed method turns out to be a generalization of some existing noise-robust ASR models, e.g., enhanced wav2vec2.0. Finally, experimental results on both synthetic and real noisy datasets show that the proposed joint training approach can improve the ASR performance under various noisy settings, leading to a stronger noise robustness.

Statistical Picking of Multivariate Waveforms

In this paper, we propose a new approach based on the fitting of a generalized linear regression model in order to detect points of change in the variance of a multivariate-covariance Gaussian variable, where the variance function is piecewise constant. By applying this new approach to multivariate waveforms, our method provides simultaneous detection of change points in functional time series. The proposed approach can be used as a new picking algorithm in order to automatically identify the arrival times of P- and S-waves in different seismograms that are recording the same seismic event. A seismogram is a record of ground motion at a measuring station as a function of time, and it typically records motions along three orthogonal axes (X, Y, and Z), with the Z-axis being perpendicular to the Earth's surface and the X- and Y-axes being parallel to the surface and generally oriented in North-South and East-West directions, respectively. The proposed method was tested on a dataset of simulated waveforms in order to capture changes in the performance according to the waveform characteristics. In an application to real seismic data, our results demonstrated the ability of the multivariate algorithm to pick the arrival times in quite noisy waveforms coming from seismic events with low magnitudes.

Precise sinusoidal signal extraction from noisy waveform in vibration calibration

Precise extraction of sinusoidal vibration parameters is essential for the dynamic calibration of vibration sensors, such as accelerometers. However, several standard methods have not yet been optimized for large background noise. In this work, signal processing methods to extract small vibration signals from noisy data in the case of accelerometer calibration are discussed. The results show that spectral leakage degrades calibration accuracy. Three methods based on the use of a filter, window function, and numerical differentiation are investigated to reduce the contribution of the calibration system noise. We demonstrate the effectiveness of these methods with theoretical calculations, simulations, and experiments. The uncertainty of microvibration calibration in the National Metrology Institute of Japan is reduced by two orders of magnitudes using the proposed methods. We recommend to use a combination of numerical differentiation and either a time-domain window function or a bandpass filter for most accelerometer microvibration sensitivity calibrations. For suppressing the effects of line noise, adjusting the data set length is also effective.

High resolution seismic waveform migration location method and its applications to induced seismicity

Locating seismic events is a central task for earthquake monitoring. Compared to arrival-based location methods, waveform-based location methods do not require picking phase arrivals and are more suitable for locating seismic events with noisy waveforms. Among waveform-based location methods, one approach is to stack different attributes of P and S waveforms around arrival times corresponding to potential event locations and origin times, and the maximum stacking values are assumed to indicate the correct event location and origin time. In this study, to obtain a high-resolution location image, we improve the waveform-based location method by applying a hybrid multiplicative imaging condition to characteristic functions of seismic waveforms. In our new stacking method, stations are divided into groups; characteristic functions of seismic waveforms recorded at stations in the same group are summed, and then multiplied among groups. We find that this approach can largely eliminate the cumulative effects of noise in the summation process and thus improve the resolution of location images. We test the new method and compare it to three other stacking methods, using both synthetic and real datasets that are related to induced seismicity occurring in petroleum/gas production. The test results confirm that the new stacking method can provide higher-resolution location images than those derived from currently used methods.

Neuronal On- and Off-type heterogeneities improve population coding of envelope signals in the presence of stimulus-induced noise

Understanding the mechanisms by which neuronal population activity gives rise to perception and behavior remains a central question in systems neuroscience. Such understanding is complicated by the fact that natural stimuli often have complex structure. Here we investigated how heterogeneities within a sensory neuron population influence the coding of a noisy stimulus waveform (i.e., the noise) and its behaviorally relevant envelope signal (i.e., the signal). We found that On- and Off-type neurons displayed more heterogeneities in their responses to the noise than in their responses to the signal. These differences in heterogeneities had important consequences when quantifying response similarity between pairs of neurons. Indeed, the larger response heterogeneity displayed by On- and Off-type neurons made their pairwise responses to the noise on average more independent than when instead considering pairs of On-type or Off-type neurons. Such relative independence allowed for better averaging out of the noise response when pooling neural activities in a mixed-type (i.e., On- and Off-type) than for same-type (i.e., only On-type or only Off-type), thereby leading to greater information transmission about the signal. Our results thus reveal a function for the combined activities of On- and Off-type neurons towards improving information transmission of envelope stimuli at the population level. Our results will likely generalize because natural stimuli across modalities are characterized by a stimulus waveform whose envelope varies independently as well as because On- and Off-type neurons are observed across systems and species.

Earthquake source dynamics and kinematics of the Eastern Indian Shield and adjoining regions

The Eastern Indian Shield (EIS) consists of two cratonic nuclei, namely Singhbhum craton and Chhotanagpur Granitic Gneissic terrain. This area contains several crisscross faults, lineaments, shear zones, numerous hot springs and three major rivers (e.g., Ganga, Brahmaputra and Damodar). The area is regionally covered by 7 seismic stations and jointly recorded 16 events from the study area, and less noisy waveforms of 4 events were used for focal mechanism analysis using the Cut and Paste method. The focal parameters of these 4 events were compiled with results of 8 events computed by different workers for the study area. To understand the detailed tectonics, focal mechanisms of 21 events for the Himalayan segment were taken from CMT Harvard catalog of duration 1976–2017. Spatial variations of operative stress fields for major tectonic domains were analyzed in the present study based on stress inversion of focal mechanism parameters. We observed strike-slip-dominated movements in the EIS, which changes partially into extension in the northeast part between the Ganga and Brahmaputra Rivers. Thrust-dominated movements evidenced by focal mechanisms and the pure compression in the western segment of the Himalaya in the north might be promoting shear movements in the EIS and adjoining regions. The pure strike-slip in the eastern segment of the Himalaya and its deeper level account for lateral shearing and eastward movements of different tectonic blocks. The normal faulting earthquakes in the northeast part might be indicating stretching in the basement because of convergence of Indian lithosphere beneath the Myanmar plate.

A model of auditory brainstem response wave I morphology

Use of the auditory brainstem response (ABR) in research has increased in the search for physiological correlates of noise-induced damage to the cochlea. The extraction of data from the ABR has traditionally relied on visual determination of peaks and troughs to calculate metrics such as wave amplitude. Visual determination can be reliable when evaluated by trained, experienced personnel, but noisy waveforms and overlapping waves produce uncertain data. The present study proposes and validates a method of fitting summed Gaussian functions to the summating potential and wave I of the ABR. This method could be useful to the research community studying these potentials by providing more accurate measures of wave amplitude than by visual determination.

The underwater vocal complexity of seals (Phocidae) is not related to their phylogeny

Closely related mammalian species often make similar vocalizations, but this is not so with the underwater calls of the true seals. Some seal species have diverse underwater vocal repertoires, whereas others only make pulsed calls. Vocal complexity scores of underwater calls of 13 seal species were compared with their phylogeny and life-history traits. Waveform types, repertoire sizes, repetition and rhythm patterns, and frequency and duration measures (15 attributes, scaled 0 to 1) were summed to give a vocal complexity score. The lowest complexity group use low frequency, burst pulse or irregular waveforms and have small repertoires. The intermediate group have both sinusoidal and noisy waveforms, songs, and a single rhythm pattern in repeated element calls. The most complex group have large repertoires, sinusoidal and noisy waveforms, songs, and two or more rhythm patterns in repeated element calls. There is no evidence of a relationship between phylogeny and vocal complexity. The low vocal complexity species are serially monogamous, do not form breeding groups, breed on beaches or pack ice, and are subject to higher predation risk. Species with higher vocal complexity are promiscuous or polygamous, form breeding groups on pack or landfast ice, and have a lower predation risk.

An algorithmic approach to single-probe Cherenkov removal in pulsed x-ray beams.

The removal of Cherenkov light in an optical dosimetry system is an important process to ensure accurate dosimetry without compromising spatial resolution. Many solutions have been presented in the literature, each with advantages and disadvantages. We present a methodology to remove Cherenkov light from a scintillator fiber optic dosimeter in a pulsed megavoltage x-ray beam using the temporal waveform across the pulse. A sample waveform of Cherenkov light can be measured by exposing only the fiber to the beam. By assuming that the Cherenkov waveform closely matches the intensity of incident radiation, this waveform can be convoluted with the instantaneous scintillation response function to generate an expected scintillation signal. By finding the least-squares fit between these two functions and the experimental data, the estimated Cherenkov contribution can be subtracted off the net signal. This can be applied for arbitrarily complex Cherenkov waveforms (within the 2ns timing resolution of the data acquisition), and in fact, the results suggest more fluctuations in the waveforms provide a better fit to data. Four beam profiles for different field sizes and energies were found with this method. They closely matched references data measured with ionization chamber with average differences across the beam no more than 4%. Noisy waveforms are assumed to be the primary cause of differences between the analyzed scintillator and IC results. We propose methods for improving the results and optimizing the data acquisition and analysis processes. These results demonstrate that it is possible and effective with a single probe to use function fitting of expected data to experimental to remove a complicated Cherenkov signal from the net light signal in pulsed-beam optical dosimetry.

Suppressing the Noise in Measured Signals for the Control of Helicopters

This paper concerns with the non-linear system having multiple-inputs multiple-outputs (MIMO). The plant mainly comprises: bench-top helicopter, tail and main rotor of a helicopter system. The dynamics are presented with control methodologies where a conventional strategy proves the instability of the system while the deadbeat and sliding mode control with linear matrix inequality regulates the future estimates. There have been disturbances like presence of unwanted ripples in the output of the non-linear systems (in case of stability also after 100[Formula: see text]s) and in the tracking of states accurately by updating the minimization error regularly. These problems originate mainly from the rotor section and are visited carefully by studying the dynamics of the blade, whereas, the design of filter makes the solution more appealing. The adaptive filter is capable of handling the frequency spectra of noise (reducing noise by 10[Formula: see text]dB), Euler angle deviations and travel angle accurately. Also, the stability analysis does not confirm the behavior in the case of bounded and a varying range of initial angular velocity. Hence, the problem of fluctuations is overcome by deadbeat and SMC-LMI approach which not only improved the ripples but also allowed the final response of the future states to be more exact and noiseless. As the previous research involved in position tracking (either translational or rotational) of these MIMO systems was concerned with software tools like MATLAB. This paper justifies its validation tested experimentally on OPAL RT hardware. The key findings involve the comparison of frequency spectra, the Euler deviation plot compared to CSL Helicopter and the three set-point variations providing accuracy in results in four modes — desired, actual, with controller-without filter and with controller-with filter. The use of adaptive filter with controllers have encouraged the suppression of noisy waveforms in the bench-top system very smoothly. The details regarding hardware setup are also discussed.

A GUI based automatic detection of seismic P-wave arrivals by using Short Term Average/Long Term Average (STA/LTA) method

Automatic detection of P wave very important to speed up the data analysis rather than manual. One of the methods used for P wave detection is Short Term Average/Long Term Average (STA/LTA). The P wave is detected automatically by STA/LTA when the STA/LTA ratio exceed from the given threshold value. This algorithm has been displayed in the form of Graphical User Interface (GUI) for make it easy to use. We have experimented the GUI to an earthquake on December 24, 2014 that recorded by ACEHSEIS network. We use four vertical components from different seismic stations and one of it has a noisy waveform. The P-wave was obtained when the STA/LTA ratio exceeded the threshold value at 0.35. We also determine the P wave with the manual detection and Pphase Picker as the comparison to our GUI of STA/LTA. The result shows that, our GUI of STA/LTA method can be used determine P wave arrivals although the waveform is noisy but the program cannot be used to determine the relatively minor phase of P wave arrivals. While the results of Pphase Picker is opposite of STA/LTA.

Micro-seismic event detection and location in underground mines by using Convolutional Neural Networks (CNN) and deep learning

Recent years have witnessed a clear trend to develop deeper and longer tunnels to meet the growing needs of mining. Micro-seismic events location is vital for predicting and avoiding the traditional mine disasters induced by high stress concentration, such as rock burst, roof caving, water inrush and slope landslide. Deep learning has become a research hotspot within the field of artificial intelligence in recent years, which has achieved significant progresses and applications in the areas of image recognition, speech recognition, language processing and computer vision. The biggest difference between the deep learning and the traditional back propagation training method is that the deep learning can automatically and independently learn the characteristics of a large amount of data without human intervention. This paper uses Convolutional Neural Network (CNN) and deep learning techniques to develop a method for identifying the Time Delay of Arrival (TDOA) and subsequently the source location of micro-seismic events in underground mines. The power spectrum and phase spectrum of cross wavelet transform calculated from the recorded seismic waves due to micro-seismic events are used as inputs to CNN. The amplitude and phase information of the cross wavelet transform power spectrum are parameters that are used without manual manipulation to build the complex mapping to predict TDOA by deep learning network. Experimental data from the in-field blast tests and simulation tests show that the proposed approach can well identify TDOA and hence detect the event source locations of the field blasting tests. It is demonstrated that the proposed approach with the CNN and deep learning techniques gives more accurate micro seismic source identifications with the recorded noisy waveforms from in-situ blast tests, as compared to several typical existing methods.

Quantification and statistical analysis of the transient visual evoked potential to a contrast-reversing pattern: A frequency-domain approach.

Visual function is often assessed by recording transient visual evoked potentials to contrast reversal of spatial patterns (tVEP-CR). This technique relies on measurements of amplitudes and peak times of a few points in the time-domain waveform, which require subjective selection of appropriate time points in a possibly noisy waveform and ignores much of the informational content in the response. Here, we introduce a set of frequency-domain measures that capture the full content of the response. Magnitude-squared coherence is used to determine the significance and reliability of magnitude measures; estimates of time delay are based on frequency-domain phase measures. In Study 1, extensive testing of a small number of observers revealed response details, and in Study 2, testing of a larger sample verified the novel frequency-domain measures and demonstrated the validity of a short-duration technique to produce reliable tVEP-CRs. In addition, Study 2 revealed adaptation effects present under prolonged stimulation conditions. Principal component analyses provided evidence for six distinct frequency mechanisms, and comparisons with time-domain measures indicated that power in high-frequency bands may be used as objective measures of excitatory input to visual cortex. A middle-frequency band captures the major peaks in the tVEP-CR waveform, and its power is highly correlated with the standard peak-to-trough amplitude measure. These novel frequency-domain indices may serve as more precise and powerful tools to assess visual function in healthy and diseased states.

Supervised learning methods for pathological arterial pulse wave differentiation: A SVM and neural networks approach

ObjectiveThe main goal of this study was to develop an automatic method based on supervised learning methods, able to distinguish healthy from pathologic arterial pulse wave (APW), and those two from noisy waveforms (non-relevant segments of the signal), from the data acquired during a clinical examination with a novel optical system. Materials and methodsThe APW dataset analysed was composed by signals acquired in a clinical environment from a total of 213 subjects, including healthy volunteers and non-healthy patients. The signals were parameterised by means of 39pulse features: morphologic, time domain statistics, cross-correlation features, wavelet features. Multiclass Support Vector Machine Recursive Feature Elimination (SVM RFE) method was used to select the most relevant features. A comparative study was performed in order to evaluate the performance of the two classifiers: Support Vector Machine (SVM) and Artificial Neural Network (ANN). Results and discussionSVM achieved a statistically significant better performance for this problem with an average accuracy of 0.9917±0.0024 and a F-Measure of 0.9925±0.0019, in comparison with ANN, which reached the values of 0.9847±0.0032 and 0.9852±0.0031 for Accuracy and F-Measure, respectively. A significant difference was observed between the performances obtained with SVM classifier using a different number of features from the original set available. ConclusionThe comparison between SVM and NN allowed reassert the higher performance of SVM. The results obtained in this study showed the potential of the proposed method to differentiate those three important signal outcomes (healthy, pathologic and noise) and to reduce bias associated with clinical diagnosis of cardiovascular disease using APW.

Chirp Signals and Noisy Waveforms for Solid-State Surveillance Radars

Since the advent of “pulse compression” radar, the “chirp” signal (Linear Frequency Modulation, LFM) has been one of the most widely used radar waveforms. It is well known that, by changing its modulation into a Non-Linear Frequency Modulation (NLFM), better performance in terms of Peak-to-Sidelobes Ratio (PSLR) can be achieved to mitigate the masking effect of nearby targets and to increase the useful dynamic range. Adding an appropriate amplitude modulation, as occurs in Hybrid-NLFM (HNLFM), the PSLR can reach very low values (e.g., PSLR < −60 dB), comparable to the two-way antenna sidelobes in azimuth. On the other hand, modern solid-state power amplifier technology, using low-power modules, requires them to be combined at the Radio Frequency (RF) stage in order to achieve the desired transmitted power. Noise Radar Technology (NRT) represents a valid alternative to deterministic waveforms. It makes use of pseudo-random waveforms—realizations of a noise process. The higher its time-bandwidth (or BT) product, the higher the (statistical) PSLR. With practical BT values, the achievable PSLR using pure random noise is generally not sufficient. Therefore, the generated pseudorandom waveforms can be “tailored” (TPW: Tailored Pseudorandom Waveforms) at will through suitable algorithms in order to achieve the desired sidelobe level, even only in a limited range interval, as shown in this work. Moreover, the needed high BT, i.e., the higher time duration T having fixed the bandwidth B, matches well with the low power solid-state amplifiers of Noise Radar. Focusing the interest on (civil) surveillance radar applications, such as ATC (Air Traffic Control) and marine radar, this paper proposes a general review of the two classes of waveforms, i.e., HNLFM and TPW.

Use of the Kalman Filter for Aortic Pressure Waveform Noise Reduction.

Clinical applications that require extraction and interpretation of physiological signals or waveforms are susceptible to corruption by noise or artifacts. Real-time hemodynamic monitoring systems are important for clinicians to assess the hemodynamic stability of surgical or intensive care patients by interpreting hemodynamic parameters generated by an analysis of aortic blood pressure (ABP) waveform measurements. Since hemodynamic parameter estimation algorithms often detect events and features from measured ABP waveforms to generate hemodynamic parameters, noise and artifacts integrated into ABP waveforms can severely distort the interpretation of hemodynamic parameters by hemodynamic algorithms. In this article, we propose the use of the Kalman filter and the 4-element Windkessel model with static parameters, arterial compliance C, peripheral resistance R, aortic impedance r, and the inertia of blood L, to represent aortic circulation for generating accurate estimations of ABP waveforms through noise and artifact reduction. Results show the Kalman filter could very effectively eliminate noise and generate a good estimation from the noisy ABP waveform based on the past state history. The power spectrum of the measured ABP waveform and the synthesized ABP waveform shows two similar harmonic frequencies.

Duty‐cycle detection for noisy arbitrary waveforms using artificial neural networks

A novel technique for detecting duty cycles of noisy arbitrary waveforms by employing artificial neural networks trained with empirical moments of asynchronously sampled waveforms’ amplitudes is presented. The proposed technique is software-based and hence can be applied flexibly to numerous waveform types without requiring any hardware changes. Furthermore, in contrast to existing duty-cycle detection methods, this technique is capable of detecting duty cycles of noisy waveforms. Results obtained for square and sawtooth waveforms demonstrate wide duty-cycle detection range of 10–95% with mean absolute percentage errors < 0.4%.

A method for unambiguous identification of on-field recorded insulator leakage current waveforms portraying electrical activity on the surface

The measurement of insulator leakage current in the field is necessary to have a better understanding of the electrical phenomena on the insulator surface. Since this electrical activity cannot be anticipated in advance, long term monitoring and storage of actual waveforms is necessary. This leads to significant accumulation of data. The presence of field noise in leakage current measurement further aggravates the problem of data size. The storage of noisy waveforms consumes available storage space without adding any significant information regarding the insulator condition. Therefore, it must be ensured that only the on-field recorded waveforms which reflect significant electrical activity on the surface are recorded. In this paper, the authors have proposed the use of Short- Time Modified Hilbert Transform (STMHT) to enhance the local characteristics of the field-recorded waveforms. It has been shown that the magnitude of the STMHT of a waveform at any instant has a good correlation with the slope of the waveform at that instant. Using this property, the proposed method is able to segregate the different types of field noise from waveforms portraying significant electrical activity. It is expected that this paper will contribute to the development of efficient devices for field monitoring of insulators.