Post-Filtering with Multichannel Power Spectral Estimation Using Joint Diagonalization in Multi-Speaker Environments

Abstract.A new method for sleep-stage classification using a causal probabilistic network as automatic classifier has been implemented and validated. The system uses features from the primary sleep signals from the brain (EEG) and the eyes (AOG) as input. From the EEG, features are derived containing spectral information which is used to classify power in the classical spectral bands, sleep spindles and K-complexes. From AOG, information on rapid eye movements is derived. Features are extracted every 2 seconds. The CPN-based sleep classifier was implemented using the HUGIN system, an application tool to handle causal probabilistic networks. The results obtained using different training approaches show agreements ranging from 68.7 to 70.7% between the system and the two experts when a pooled agreement is computed over the six subjects. As a comparison, the interrater agreement between the two experts was found to be 71.4%, measured also over the six subjects.

Download Full-text

Enhancing the quality of communication of cellular networks using big data applications

Journal Of Big Data ◽

10.1186/s40537-021-00533-8 ◽

2021 ◽

Vol 8 (1) ◽

Author(s):

Mozamel Musa Saeed ◽

Mohammed Alsharidah

Keyword(s):

Big Data ◽

Spectral Density ◽

Cellular Networks ◽

Power Spectral Density ◽

Spectral Estimation ◽

Estimation Methods ◽

Spectrum Estimation ◽

Spectral Densities ◽

Power Spectral ◽

Welch Method

AbstractBoth software-defined networking and big data have gained approval and preferences from both industry and academia. These two important realms have conventionally been addressed independently in wireless cellular networks. The discussion taken into consideration in this study was to analyze the wireless cellular technologies with the contrast of efficient and enhanced spectral densities at a reduced cost. To accomplish the goal of this study, Welch's method has been used as the core subject. With the aid of previous research and classical techniques, this study has identified that the spectral densities can be enhanced at reduced costs with the help of the power spectral estimation methods. The Welch method gives the result on power spectrum estimation. By reducing the effect of noise, the Welch method is used to calculate the power spectral density of a signal. When data length is increased, Welch's method is considered the best as a conclusion to this paper because excellent results are yielded by it in the area of power spectral density estimation.

Download Full-text

Robust Feature Vector Set Using Higher Order Autocorrelation Coefficients

Developments in Natural Intelligence Research and Knowledge Engineering ◽

10.4018/978-1-4666-1743-8.ch009 ◽

2012 ◽

pp. 126-134

Author(s):

Poonam Bansal ◽

Amita Dev ◽

Shail Jain

Keyword(s):

Speech Recognition ◽

Speech Signal ◽

Background Noise ◽

Extraction Method ◽

Recognition Performance ◽

Spectral Estimation ◽

Higher Order ◽

Feature Extraction Method ◽

Power Spectral ◽

Cepstral Coefficients

In this paper, a feature extraction method that is robust to additive background noise is proposed for automatic speech recognition. Since the background noise corrupts the autocorrelation coefficients of the speech signal mostly at the lower orders, while the higher-order autocorrelation coefficients are least affected, this method discards the lower order autocorrelation coefficients and uses only the higher-order autocorrelation coefficients for spectral estimation. The magnitude spectrum of the windowed higher-order autocorrelation sequence is used here as an estimate of the power spectrum of the speech signal. This power spectral estimate is processed further by the Mel filter bank; a log operation and the discrete cosine transform to get the cepstral coefficients. These cepstral coefficients are referred to as the Differentiated Relative Higher Order Autocorrelation Coefficient Sequence Spectrum (DRHOASS). The authors evaluate the speech recognition performance of the DRHOASS features and show that they perform as well as the MFCC features for clean speech and their recognition performance is better than the MFCC features for noisy speech.

Download Full-text

Power Spectral Estimation

Introduction to Digital Signal Processing ◽

10.1016/b978-0-12-398420-3.50015-1 ◽

1989 ◽

pp. 258-291

Author(s):

John H. Karl

Keyword(s):

Spectral Estimation ◽

Power Spectral

Download Full-text

Resolving the Issues of Capon and APES Approach for Projecting Enhanced Spectral Estimation

International Journal of Electrical and Computer Engineering (IJECE) ◽

10.11591/ijece.v6i2.9034 ◽

2016 ◽

Vol 6 (2) ◽

pp. 725

Author(s):

Kantipudi MVV Prasad ◽

H.N. Suresh

Keyword(s):

Signal Processing ◽

Spectral Analysis ◽

Spectral Density ◽

Power Spectral Density ◽

Simple Technique ◽

Spectral Estimation ◽

Research Work ◽

Research Community ◽

The Past ◽

Power Spectral

There are various applications on signal processing that is highly dependent on preciseness and accuracy of the outcomes in spectrum of signals. Hence, from the past two decades the research community has recognized the benefits, significance, as well as associated problems in carrying out a model for spectral estimation. While in-depth investigation of the existing literatures shows that there are various attempts by the researchers to solve the issues associated with spectral estimations, where majority of teh research work is inclined towards addressing problems associated with Capon and APES techniques of spectral analysis. Therefore, this paper introduces a very simple technique towards resolving the issues of Capon and APES techniques. The outcome of the study was analyzed using correlational factor and power spectral density to find the proposed system offers better spectral estimations compared to existing system.

Download Full-text

A study on wavelet analysis of SSVEP Signals

International Journal of Engineering & Technology ◽

10.14419/ijet.v7i2.25.12354 ◽

2018 ◽

Vol 7 (2.25) ◽

pp. 10

Author(s):

Bincy Babu ◽

R Chandrasekaran ◽

Josline Elsa Joseph ◽

Thella Shalem Rahul ◽

T R Thamizhvani ◽

...

Keyword(s):

Wavelet Analysis ◽

Visual Evoked Potential ◽

Evoked Potential ◽

Spectral Estimation ◽

Power Spectral Analysis ◽

Feature Analysis ◽

Eeg Signals ◽

Power Spectral ◽

Frequency Components ◽

Bci System

Almost every Brain Control Interfcae (BCI) system is framed based on Steady State Visual Evoked Potential (SSVEP) which is predicted through distinguishing overriding frequency components in Electroencephalography (EEG) signals. The proposed system aims in accurate feature extraction of SSVEP signals. Power spectral analysis and wavelet analysis are done for feature analysis. The feature set variation for male and female subjects are obtained. Compared power spectral estimation and wavelet analysis, merits and demerits of each approach can be identified from the outcomes. It offers a theoretical reference of practical choice for BCI application.

Download Full-text

Spectral Estimation of Nasal Cyclic Rhythm by Nasal Airflow Temperature Measurement

Applied Mechanics and Materials ◽

10.4028/www.scientific.net/amm.573.848 ◽

2014 ◽

Vol 573 ◽

pp. 848-855

Author(s):

R. Sukanesh ◽

E. Muthu Kumaran

Keyword(s):

Pathological Condition ◽

Spectral Estimation ◽

Power Spectral Analysis ◽

Nasal Airflow ◽

Airway Patency ◽

Nasal Cycle ◽

Power Spectral ◽

Left And Right ◽

The Difference ◽

Cyclic Fluctuation

.The nasal cycle is referred to a cyclic fluctuation in congestion of the nasal mucosa that results in rhythmic and bilateral reciprocal alteration of nasal airway patency. The purpose of this study is to deal with statistical and power spectral analysis of nasal cycle by measuring the temperature difference between the airflow of both left and right nostrils. Five adult voluntary healthy subjects are enrolled for the study. Nasal temperature probe combined with amplifier are used for recording nasal airflow temperature on both nostrils. The highest nasal airflow temperature values are detected at the end of expiration and the lowest values are detected at the end of inspiration. Nasal cycle found in all the subjects and lasted to the minimum of 30 minutes to maximum of 6 hours. The difference in temperature of both nostrils is statistically significant (p<0.05) and spectral estimation is made using autoregressive modeling. The method is used to quantify nasal obstruction in pathological condition and also to correlate the related physiological phenomenon.

Download Full-text