Enhancement of Maternal ECG Using Short-Term Fourier Transform for Foetal Electrocardiogram Extraction

The objective of the project is to improve maternal abdomen recording for better prediction of foetal Electrocardiogram (FECG). One of the most difficult tasks in observing foetal well-being is obtaining a clean foetal Electrocardiogram (FECG) using non-invasive abdominal recordings. The foetal graph's low signal quality, on the other hand, makes morphological examination of its wave structure in clinical follow-up difficult. The signal contains precise information that can help doctors to monitor fetal health during pregnancy and labor. The abdominal signal is normalized and separated in the pre-processing stage for wave shape analysis in clinical follow-up. The Kaiser window is used for spectral analysis and segmenting the signal. The two-dimensional (2D) time-frequency representation is obtained by short-time Fourier transform (STFT). The STFT enhances the abdominal recordings of maternal Electrocardiogram (MECG) for efficient separation of foetal electrocardiogram (FECG) to monitor the foetus well-being.

Through-the-Wall Micro-Doppler De-Wiring Technique via Cycle-Consistent Adversarial Network

The radar penetrating technique has aroused a keen interest in the research community, due to its superior abilities for through-the-wall indoor human motion monitoring. Micro-Doppler signatures in this situation play a significant role in recognition and classification for human activities. However, the live wire buried in the wall introduces additive clutters to the spectrograms. Such degraded spectrograms drastically affect the performance of behind-the-wall human activity detection. In this paper, an ultra-wideband (UWB) radar system is utilized in the through-the-wall scenario to get the feature enhanced micro-Doppler signature called range-max time-frequency representation (R-max TFR). Then, a recently introduced Cycle-Consistent Generative Adversarial Network (Cycle GAN) is employed to realize the end-to-end de-wiring task. Cycle GAN can learn the mapping between spectrograms with and without the live wire effect. To minimize the wiring clutters, a loss function called identity loss is introduced in this work. Finally, the proposed de-wiring approach is evaluated through classification. The results show that the proposed Cycle GAN architecture outperforms other state-of-art de-wiring methods.

Optimization and software-numerical implementation of miller's algorithm on a four-frequency model of solid vibration movement

On the basis of a programmed-numerical approach, new values of the coefficients in the Miller orientation algorithm are obtained. For this, an analytical reference model of the angular motion of a rigid body was applied in the form of a four-frequency representation of the orientation quaternion.The numerical implementation of the reference model for a given set of frequencies is presented in the form of constructed trajectories in the configuration space of orientation parameters. A software-numerical implementation of Miller's algorithm is carried out for different values of the coefficients and the values of the coefficients are obtained, which optimize the error of the accumulated drift. It is shown that for the presented reference model of angular motion, Miller's algorithm with a new set of coefficients provides a lower computational drift error compared to with the classic Miller algorithm and the Ignagni modification, which are optimized for conical motion.

Minimally-modeled search of higher multipole gravitational-wave radiation in compact binary coalescences.

As the Advanced LIGO and Advanced Virgo interferometers, soon to be joined by the KAGRA interferometer, increase their sensitivity, they detect an ever-larger number of gravitational waves with a significant presence of higher multipoles in addition to the dominant (2, 2) multipole. These higher multipoles can be detected with different approaches, such as the minimally-modeled burst search methods, and here we discuss one such approach based on the coherent WaveBurst pipeline (cWB). During the inspiral phase the higher multipoles produce chirps whose instantaneous frequency is a multiple of the dominant (2, 2) multipole, and here we describe how cWB can be used to detect these spectral features. The search is performed within suitable regions of the time-frequency representation; their shape is determined by optimizing the Receiver Operating Characteristics. This novel method has already been used in the GW190814 discovery paper (Astrophys. J. Lett. 896 L44) and is very fast and flexible. Here we describe in full detail the procedure used to detect the (3, 3) multipole in GW190814 as well as searches for other higher multipoles during the inspiral phase, and apply it to another event that displays higher multipoles, GW190412, replicating the results obtained with different methods. The procedure described here can be used for the fast analysis of higher multipoles and to support the findings obtained with the model-based Bayesian parameter estimates.

Unsupervised Learning Architecture for Classifying the Transient Noise of Interferometric Gravitational-Wave Detectors

In the data of laser interferometric gravitational wave detectors, transient noise with non-stationary and non-Gaussian features occurs at a high rate. It often causes problems such as instability of the detector, hiding and/or imitating gravitational-wave signals. This transient noise has various characteristics in the time-frequency representation, which is considered to be associated with environmental and instrumental origins. Classification of transient noise can offer one of the clues for exploring its origin and improving the performance of the detector. One approach for the classification of these noises is supervised learning. However, generally, supervised learning requires annotation of the training data, and there are issues with ensuring objectivity in the classification and its corresponding new classes. On the contrary, unsupervised learning can reduce the annotation work for the training data and ensuring objectivity in the classification and its corresponding new classes. In this study, we propose an architecture for the classification of transient noise by using unsupervised learning, which combines a variational autoencoder and invariant information clustering. To evaluate the effectiveness of the proposed architecture, we used the dataset (time-frequency two-dimensional spectrogram images and labels) of the LIGO first observation run prepared by the Gravity Spy project. We obtain the consistency between the label annotated by Gravity spy project and the class provided by our proposed unsupervised learning architecture and provide the potential for the existence of the unrevealed classes.

K-nearest neighbor and naïve Bayes based diagnostic analytic of harmonic source identification

This paper proposes a comparison of machine learning (ML) algorithm known as the k-nearest neighbor (KNN) and naïve Bayes (NB) in identifying and diagnosing the harmonic sources in the power system. A single-point measurement is applied in this proposed method, and using the S-transform the measurement signals are analyzed and extracted into voltage and current parameters. The voltage and current features that estimated from time-frequency representation (TFR) of S-transform analysis are used as the input for MLs. Four significant cases of harmonic source location are considered, whereas harmonic voltage (HV) and harmonic current (HC) source type-load are used in the diagnosing process. To identify the best ML, the performance measurement of the proposed method including the accuracy, precision, specificity, sensitivity, and F-measure are calculated. The sufficiency of the proposed methodology is tested and verified on IEEE 4-bust test feeder and each ML algorithm is executed for 10 times due to prevent any overfitting result.

Sinusoidal Noise Removal in PD Measurement Based on Synchrosqueezing Transform and Singular Spectrum Analysis

In electrical engineering, partial discharge (PD) measurement has been widely used for inspecting and judging insulation conditions of high voltage (HV) apparatus. However, on-site PD measurement easily becomes contaminated by noises. Particularly, sinusoidal noise makes it difficult to recognize real PD signal, thus leading to the misjudgment of insulation conditions. Therefore, sinusoidal noise removal is necessary. In this paper, instantaneous frequency (IF) is introduced, and the synchrosqueezing transform (SST) as well as singular spectrum analysis (SSA) is proposed for sinusoidal noise removal. A continuous analytic wavelet transform is firstly applied to the noisy PD signal and then the time frequency representation (TFR) is reassigned by SST. Narrow-band sinusoidal noise has fixed IF, while PD signal has much larger frequency range and time-varying IF. Due to the difference, the reassigned TFR enables the sinusoidal noise to be distinguished from PD signal. After synthesizing the signal with the recognized IF, SSA is further applied to signal refinement. At last, a numerical simulation is carried out to verify the effectiveness of the proposed method, and its robustness to white noise is also validated. After the implementation of the proposed method, wavelet thresholding can be further applied for white noise reduction.