Linear Predictive Cepstrum Research Articles

A Stator Slot Wedge Loosening Offline Detection System Based on an Intelligent Maintenance Robot of a Large Hydro Generator

Hydropower is at the backbone of low-carbon clean energy, while large hydroelectric generators play a key role in hydropower systems. In this paper, the tightness detection of stator trough wedges in the offline state of large hydroelectric generators is mainly studied. The offline state means that the generator stops running but does not need to move the stator and rotor. The traditional detection of generator stator wedges has problems, such as long maintenance intervals and low work efficiency. According to the structural characteristics of the generator, a generator maintenance robot device based on the track mechanism is designed. The device can simultaneously visualize the internal of the generator and detect the tightness of the stator slot wedge, effectively improving the maintenance efficiency. According to the electromagnetic magnitude of the stator rod in the alternating magnetic field, different stator slot wedge models are built for tightening, slightly tightening and loosening. For the conventional slot wedge loosen detection, there is a problem that the characteristic parameter is single, and the state of the slot wedge cannot be fully fed back. In this paper, a method for extracting the Linear Prediction Cepstrum Coefficient (LPCC) and Mel Frequency Cepstrum Coefficient (MFCC) of percussion sound signal is proposed, and the tightness recognition of the stator slot wedge is realized by combining the BP neural network algorithm. Experimental results show that the proposed method can effectively identify stator slot wedges in different states.

Recognition of behavior state of Penaeus vannamei based on passive acoustic technology

Underwater signal processing based on passive acoustic technology has carried out a lot of research on the behavioral sound of underwater creatures and the protection of marine resources, which proves the feasibility of passive acoustic technology for long-term and non-destructive monitoring of underwater biological sound production. However, at present, most relevant research focuses on fish but little research on shrimp. At the same time, as the main economic breeding industry, Penaeus vannamei has a backward industrial structure, in which the level of intelligence needs to be improved. In this paper, the acoustic signals generated by different physiological behaviors of P. vannamei are collected based on passive acoustic technology. Their different behaviors are finally classified and identified through feature extraction and analysis. Meanwhile, the characteristic non-parametric ANOVA is carried out to explore the relationship between the acoustic signals and the behavior state of P. vannamei to achieve the purpose of real-time monitoring of the behavior state of P. vannamei. The experimental results show that linear prediction cepstrum coefficient (LPCC) and Mel-frequency cepstrum coefficient (MFCC) characteristic coefficients are effective in the classification and recognition of different behavioral acoustic signals with interspecific acoustic signals of P. vannamei. Meanwhile, the SVM classifier based on OvR classification strategy can model the acoustic signal characteristics of different underwater biological behaviors more efficiently and has classification accuracy as high as 93%.

Robust cepstral feature for bird sound classification

Birds are excellent environmental indicators and may indicate sustainability of the ecosystem; birds may be used to provide provisioning, regulating, and supporting services. Therefore, birdlife conservation-related researches always receive centre stage. Due to the airborne nature of birds and the dense nature of the tropical forest, bird identifications through audio may be a better solution than visual identification. The goal of this study is to find the most appropriate cepstral features that can be used to classify bird sounds more accurately. Fifteen (15) endemic Bornean bird sounds have been selected and segmented using an automated energy-based algorithm. Three (3) types of cepstral features are extracted; linear prediction cepstrum coefficients (LPCC), mel frequency cepstral coefficients (MFCC), gammatone frequency cepstral coefficients (GTCC), and used separately for classification purposes using support vector machine (SVM). Through comparison between their prediction results, it has been demonstrated that model utilising GTCC features, with 93.3% accuracy, outperforms models utilising MFCC and LPCC features. This demonstrates the robustness of GTCC for bird sounds classification. The result is significant for the advancement of bird sound classification research, which has been shown to have many applications such as in eco-tourism and wildlife management.

Application of wearable motion sensor in business English teaching

With the advancement of science and technology, portable motion sensors are becoming increasingly popular in life and have become a research point for improving life and learning, and are widely used in medical care and smart terminals. Based on the advantages of portable motion sensors, this paper focuses on its application in learning English business. Collects speech information through special motion sensors, analyzes the accuracy of students' reading through speech recognition, to help students better learn business English. Firstly, the wearable sensor is used to collect and preprocess the speech information of students' business English reading as the input of speech recognition. Secondly, the linear predictive cepstrum coefficient (LPCC) and Meier frequency cepstrum coefficient (MFCC) of students' business English reading speech are extracted, and the mixed parameters of LPCC and MFCC are taken as speech features. Finally, the correctness of reading speech is recognized by combining HMM and WNN. Through the simulation analysis of students' reading speech recognition, it is shown that the speech recognition based on wearable motion sensor is feasible and the recognition method has good performance. In addition, the feasibility of wearable motion sensors in business English teaching is verified by the establishment of an experimental classes, which can promote students' English learning better.

Classification between Elderly Voices and Young Voices Using an Efficient Combination of Deep Learning Classifiers and Various Parameters

The objective of this research was to develop deep learning classifiers and various parameters that provide an accurate and objective system for classifying elderly and young voice signals. This work focused on deep learning methods, such as feedforward neural network (FNN) and convolutional neural network (CNN), for the detection of elderly voice signals using mel-frequency cepstral coefficients (MFCCs) and linear prediction cepstrum coefficients (LPCCs), skewness, as well as kurtosis parameters. In total, 126 subjects (63 elderly and 63 young) were obtained from the Saarbruecken voice database. The highest performance of 93.75% appeared when the skewness was added to the MFCC and MFCC delta parameters, although the fusion of the skewness and kurtosis parameters had a positive effect on the overall accuracy of the classification. The results of this study also revealed that the performance of FNN was higher than that of CNN. Most parameters estimated from male data samples demonstrated good performance in terms of gender. Rather than using mixed female and male data, this work recommends the development of separate systems that represent the best performance through each optimized parameter using data from independent male and female samples.

Analysis and classification of speech sounds of children with autism spectrum disorder using acoustic features

Children with autism spectrum disorder (ASD) produce speech sounds different from that of Normal or non-ASD children. Hence, analyzing acoustic features can help characterizing the ASD speech signals. In this study, the distinguishing characteristics of speech production are examined for ASD affected children, with comparison to Normal children’s speech. Acoustic features are analyzed first and then classification of ASD vs Normal speech is attempted using different machine learning techniques. Two speech sound databases are recorded for this study: the speech database of children with ASD and the speech database of Normal children. English speech utterances are recorded for children of Indian regional (Tamil and Telugu) nativity. The changes due to autism effect are examined in context of 5 English vowel sounds (/a/, /e/, /i/, /o/, and /u/). Changes in the speech production characteristics of children are explored using three sets of features. Firstly, changes in the excitation source features are examined using strength of excitation (SoE) and instantaneous fundamental frequency (F0). Secondly, changes in the vocal tract (VT) filter features are examined using dominant frequencies (FD1, FD2) and formant frequencies (F1 to F5). Thirdly, changes in the source-filter combined features are examined using signal energy (E), zero-crossing rate (ZCR), linear prediction cepstrum coefficients (LPCC), and Mel-frequency cepstral coefficients (MFCC). Then, various combinations of the acoustic features are classified utilizing machine learning methods such as probabilistic neural network (PNN), multilayer perceptron (MLP), support vector machine (SVM), and K-nearest neighbors (KNN). Analyses of acoustic features shows significant differences between the speech of children with ASD and the Normal children. Results up to 98.17% accuracy are obtained for classification between acoustic features of the speech sounds of children with ASD and the Normal children. The observations and this study results may be useful as acoustic biomarkers to identify autism and its progression/cure among children. This study may also be valuable towards developing a system for ASD diagnosis from children’s speech sounds, in the future.

Automatic Speech Emotion Recognition Using Machine Learning

For several years, emotion detection from speech signals has been a research topic in human-machine interface applications. To discern emotions from speech signals, a variety of devices have been developed. Theoretical definitions, categorizations, and modalities of emotion expression are all discussed. To conduct this research, a SER framework based on various classifiers and feature extraction methods was developed. The mel-frequency cepstrum coefficients (MFCC) and modulation spectral (MS) characteristics of speech signals are analysed and fed into various classifiers for training. Using feature selection, this method is used to find the most important function subset (FS). The features extracted from emotional speech samples that make up the database for the speech emotion recognition system include power, pitch, linear prediction cepstrum coefficient (LPCC), and Mel frequency cepstrum coefficient (MFCC). The effectiveness of classification is determined by the extracted features. Seven emotions are classified using a recurrent neural network (RNN) classifier. Their results are then compared to techniques such as multivariate linear regression (MLR) and support vector machines that are used in the field of emotion detection for spoken audio signals (SVM).

Experimental Evaluation of Deep Learning Methods for an Intelligent Pathological Voice Detection System Using the Saarbruecken Voice Database

This work is focused on deep learning methods, such as feedforward neural network (FNN) and convolutional neural network (CNN), for pathological voice detection using mel-frequency cepstral coefficients (MFCCs), linear prediction cepstrum coefficients (LPCCs), and higher-order statistics (HOSs) parameters. In total, 518 voice data samples were obtained from the publicly available Saarbruecken voice database (SVD), comprising recordings of 259 healthy and 259 pathological women and men, respectively, and using /a/, /i/, and /u/ vowels at normal pitch. Significant differences were observed between the normal and the pathological voice signals for normalized skewness (p = 0.000) and kurtosis (p = 0.000), except for normalized kurtosis (p = 0.051) that was estimated in the /u/ samples in women. These parameters are useful and meaningful for classifying pathological voice signals. The highest accuracy, 82.69%, was achieved by the CNN classifier with the LPCCs parameter in the /u/ vowel in men. The second-best performance, 80.77%, was obtained with a combination of the FNN classifier, MFCCs, and HOSs for the /i/ vowel samples in women. There was merit in combining the acoustic measures with HOS parameters for better characterization in terms of accuracy. The combination of various parameters and deep learning methods was also useful for distinguishing normal from pathological voices.

Real-time Monitoring of Fluidized Bed Agglomerating based on Improved Adaboost Algorithm

Absctract: In order to to detect the polymer agglomeration in fluidized bed reactor (FBR), a method of real-time monitoring of agglomeration in fluidized bed polyolefin reactor based on voiceprint feature recognition is developed. First, the acoustic emission detection technology is applied to collect the acoustic signal generated by the polymer collision on the inner wall of FBR. Then, the voiceprint features of the collected acoustic signal are extracted with the Mel Frequency Cepstrum Coefficients (MFCC) and the Linear Prediction Cepstrum Coefficients (LPCC). To classify the extracted voiceprint features, an improved Adaboost algorithm is proposed to establish the real-time agglomeration classification model. Due to the introduction of cost factor and Gini index decision-making calculation to the Adaboost algorithm, the proposed improved Adaboost algorithm can classify unbalanced small samples with better accuracy and F-score index compared with the traditional Adaboost algorithm. The experiment results in a fluidized bed pilot plant have verified the effectiveness and feasibility of the proposed method.

Cobalt crust recognition based on kernel Fisher discriminant analysis and genetic algorithm in reverberation environment

Recognition of substrates in cobalt crust mining areas can improve mining efficiency. Aiming at the problem of unsatisfactory performance of using single feature to recognize the seabed material of the cobalt crust mining area, a method based on multiple-feature sets is proposed. Features of the target echoes are extracted by linear prediction method and wavelet analysis methods, and the linear prediction coefficient and linear prediction cepstrum coefficient are also extracted. Meanwhile, the characteristic matrices of modulus maxima, sub-band energy and multi-resolution singular spectrum entropy are obtained, respectively. The resulting features are subsequently compressed by kernel Fisher discriminant analysis (KFDA), the output features are selected using genetic algorithm (GA) to obtain optimal feature subsets, and recognition results of classifier are chosen as genetic fitness function. The advantages of this method are that it can describe the signal features more comprehensively and select the favorable features and remove the redundant features to the greatest extent. The experimental results show the better performance of the proposed method in comparison with only using KFDA or GA.

A Unique Glottal Flow Parameters based Features for Anti-spoofing Countermeasures in Automatic Speaker Verification

The domain of Automatic Speaker Verification (ASV) is blooming with growing developments in feature en-gineering and artificial intelligence. Inspite of this, the system is liable to spoofing attacks in the form of synthetic or replayed speech. The difficulty in detecting synthetic speech is due to recent advancements in the Voice conversion and Text-to-speech systems which produce natural, indistinguishable speech. To prevent such attacks, there is a need to develop robust spoof detection systems. In order to achieve this goal, we are proposing estimation of Glottal Flow Parameters (GFP) from speech of genuine speech and synthetic spoof samples. The GFP are further parameterized using time, frequency and Liljencrants–Fant (LF) models. Along with GFP features, the Linear Prediction Cepstrum Co-efficient (LFCC) and statistical parameters are computed. The GFP features are investigated to prove their usefulness in detecting spoofed and genuine speech. The ASV spoof 2019 corpus is used to test the framework and evaluated against the baseline models. The proposed spoof detection framework produces an Equal Error Rate (EER) of 2.39% and tandem Detection Cost Function (t-DCF) of 0.0562 which is found to be better than the state-of-the art technique.

Comparison of Khasi speech representations with different spectral features and hidden Markov states

In this paper, we present a comparison of the Khasi speech representations with four different spectral features and novel extension towards the development of Khasi speech corpora. These four features include linear predictive coding (LPC), linear prediction cepstrum coefficient (LPCC), perceptual linear prediction (PLP), and Mel frequency cepstral coefficient (MFCC). The 10-h speech data was used for training and 3-h data for testing. For each spectral feature, different hidden Markov model (HMM) based recognizers with variations in HMM states and different Gaussian mixture models (GMMs) were built. The performance was evaluated by using the word error rate (WER). The experimental results showed that MFCC provides a better representation for Khasi speech compared with the other three spectral features.

Speech stress recognition using semi-eager learning

Homo-sapiens suffer from psychogenic pain due to current day lifestyle. According to psychologists, stress is the most destructive form of psychalgia and it is a vicious companion for this species. Immoderate levels of stress may lead to the death of many individuals. Normally, the presence of stress gives rise to certain emotions which can be detected to predict stress levels of a person. This paper proposes the development of mechanized and efficient Speech Emotion Recognition (SER) for stress level analysis. The paper investigates the performance of perceptual based speech features like Revised Perceptual Linear Prediction Coefficients, Bark Frequency Cepstral Coefficients, Perceptual Linear Predictive Cepstrum, Gammatone Frequency Cepstral coefficient, Mel Frequency Cepstral Coefficient, Gammatone Wavelet Cepstral Coefficient and Inverted Mel Frequency Cepstral Coefficients on SER. The novelty of this work involves application of a SemiEager (SemiE) learning algorithm for evaluating auditory cues. SemiE offers advantages over eager and lazy based learning by reducing the computational cost. Stress level recognition being the main objective, the Speech Under Simulated and Actual Stress (SUSAS) benchmark database is used for performance analysis. A comparative analysis is presented to demonstrate the improvement in the SED performance. An overall accuracy of 90.66% recognition of stress related emotions is achieved.

Automatic Recognition of Arabic Poetry Meter from Speech Signal using Long Short-term Memory and Support Vector Machine

The recognition of the poetry meter in spoken lines is a natural language processing application that aims to identify a stressed and unstressed syllabic pattern in a line of a poem. Stateof-the-art studies include few works on the automatic recognition of Arud meters, all of which are text-based models, and none is voice based. Poetry meter recognition is not easy for an ordinary reader, it is very difficult for the listener and it is usually performed manually by experts. This paper proposes a model to detect the poetry meter from a single spoken line (“Bayt”) of an Arabic poem. Data of 230 samples collected from 10 poems of Arabic poetry, including three meters read by two speakers, are used in this work. The work adopts the extraction of linear prediction cepstrum coefficient and Mel frequency cepstral coefficient (MFCC) features, as a time series input to the proposed long short-term memory (LSTM) classifier, in addition to a global feature set that is computed using some statistics of the features across all of the frames to feed the support vector machine (SVM) classifier. The results show that the SVM model achieves the highest accuracy in the speakerdependent approach. It improves results by 3%, as compared to the state-of-the-art studies, whereas for the speaker-independent approach, the MFCC feature using LSTM exceeds the other proposed models.

Khasi Speech Recognition using Hidden Markov Model with Different Spectral Features: A Comparison

In this work, various feature extraction techniques are implemented for a speech data corpus in Khasi Language. This paper provides a comparative performance analysis of widely used spectral features like Mel-Frequency Cepstrum Coefficients (MFCC), Perceptual Linear Prediction (PLP), Linear Predictive Cepstrum Coefficient (LPCC), Linear Predictive Coefficient (LPC) and Linear Predictive Reflection Coefficients (LPREFC). For each features we have used coefficient dimensions of static, delta and acceleration evaluated for the recognition of words spoken in Khasi language, a language spoken in the Khasi and Jaintia hills districts of Meghalaya. A recorder based continuous speech database has been collected as per the standard procedure and data processing was done using wave surfer. From the analysis, it was observed that for sampling rate of 16 KHz and 8 KHz the word recognition accuracy of MFCC and PLP features are better than LPCC, LPC and LPREFC features under all circumstances.

Separating Multi-Source Partial Discharge Signals Using Linear Prediction Analysis and Isolation Forest Algorithm

Partial discharge (PD) detection is an effective way to find defects and diagnose the insulation condition of power equipment. During manufacturing and operating, there may exist multi-source PD signals in the equipment, which could seriously affect the accuracy of subsequent defect analyses. In this paper, a universal separation methodology using linear prediction analysis (LPA) and the isolation forest algorithm (IFA) is introduced. By approximating the present waveform point by a linear combination of several points in the past, a 12-D linear prediction cepstrum coefficient (LPCC) feature space can be established which can accurately characterize PD waveforms. In this paper, we applied principal component analysis (PCA) to reduce the feature space to 2-D space. The IFA was adopted to separate multi-source PD signals which quantified the degree of clustering and added one more parameter based on the original features. Thus, the proposed method can separate multiple PD sources effectively even if the difference in the features of different types of PDs is small. In addition, the proposed algorithm easily removed the noise points during the separation process. The algorithm was applied to different types of PD signals on 35-kV transformers and a three-source PD data set on a 252-kV gas-insulated switchgear (GIS) platform. The separation results confirm that the proposed algorithm can effectively separate and distinguish various PD signals.

A Pathological Multi-Vowels Recognition Algorithm Based on LSP Feature

At present, pathological voice recognition is mainly based on the classification of pathological voice. However, almost all the researches are based on the single vowel \a samples, but few on multi-vowels. In addition, the current researches on multi-vowels recognition are mainly for normal voices, which are unsuitable for the speech recognition of normal and pathological multi-vowels simultaneously. This paper concentrates on developing an accurate and robust feature called enhanced-bark line spectrum pair (E-BLSP) to detect and classify normal and pathological multi-vowels. We explore the impact of E-BLSP feature on recognition performance and propose an effective method based on the combination of three features including E-BLSP for pathological and normal multi-vowels. In this paper, first LSP and difference of adjacent LSP (DAL) features of a vowel are extracted. Then LSP feature is warped at bark domain to get bark line spectrum pair (BLSP). In addition, then E-BLSP feature is calculated by adjusting BLSP using DAL feature. Finally, the adjusted E-BLSP feature and other two traditional features, including linear prediction cepstrum coefficient (LPCC) and mel-frequency cepstrum coefficients (MFCC) are applied to support vector machine (SVM) and deep neural network (DNN) classifiers to explore the classification performance of single feature and feature combinations for pathological and normal vowels /a/, /i/ and /u/. The results show that the highest achieved accuracies for DNN and SVM network are 98.6190% and 96.2693%, while the largest achieved area under curves (AUC) are 0.9925 and 0.9868, correspondingly with the combination of three features including LPCC, MFCC, and E-BLSP.

Enhanced speech emotion detection using deep neural networks

This paper focusses on investigation of the effective performance of perceptual based speech features on emotion detection. Mel frequency cepstral coefficients (MFCC’s), perceptual linear predictive cepstrum (PLPC), Mel frequency perceptual linear prediction cepstrum (MFPLPC), bark frequency cepstral coefficients (BFCC), revised perceptual linear prediction coefficient’s (RPLP) and inverted Mel frequency cepstral coefficients (IMFCC) are the perception features considered. The algorithm using these auditory cues is evaluated with deep neural networks (DNN). The novelty of the work involves analysis of the perceptual features to identify predominant features that contain significant emotional information about the speaker. The validity of the algorithm is analysed on publicly available Berlin database with seven emotions in 1-dimensional space termed categorical and 2-dimensional continuous space consisting of emotions in valence and arousal dimensions. Comparative analysis reveals that considerable improvement in the performance of emotion recognition is obtained using DNN with the identified combination of perceptual features.

Emotions recognition: different sets of features and models

The better and effective human–machine communication is ensured by performing affective computing. In the recent years, healthy research has been progressing in recognizing emotions by using various databases. This paper mainly emphasizes the effectiveness on the basis of using different sets of features and modeling techniques in evaluating the performance of multiple speaker-independent and speaker-dependent emotion recognition systems. It has become a challenging task to improve the performance of emotion recognition system, since EMO-DB Berlin database used in this work contains only ten speeches uttered by ten speakers in different emotions namely, Anger, Boredom, Disgust, Fear, Happiness, Sadness and Neutral. Speaker dependent and independent emotion recognition is done by creating models using clustering technique, Gaussian mixture modeling (GMM) and continuous density hidden Markov modeling (CDHMM) techniques for all emotions. The emotion recognition system is also evaluated for mel frequency cepstrum (MFCC) and concatenated MFCC with probability & shifted delta cepstrum (SDC), mel frequency linear predictive cepstrum (MFPLPC) and concatenated MFPLPC with probability & SDC and formants for clustering used as a modeling technique. These features provide complementary evidence in assessing the performance of the system based on VQ based clustering technique. This algorithm provides 99 and 100% as overall weighted accuracy recall (WAR) for performance evaluation with respect to correct identification of emotion for any one feature and modeling technique.

Use of Acoustic Emission and Pattern Recognition for Crack Detection of a Large Carbide Anvil.

Large-volume cubic high-pressure apparatus is commonly used to produce synthetic diamond. Due to the high pressure, high temperature and alternative stresses in practical production, cracks often occur in the carbide anvil, thereby resulting in significant economic losses or even casualties. Conventional methods are unsuitable for crack detection of the carbide anvil. This paper is concerned with acoustic emission-based crack detection of carbide anvils, regarded as a pattern recognition problem; this is achieved using a microphone, with methods including sound pulse detection, feature extraction, feature optimization and classifier design. Through analyzing the characteristics of background noise, the cracked sound pulses are separated accurately from the originally continuous signal. Subsequently, three different kinds of features including a zero-crossing rate, sound pressure levels, and linear prediction cepstrum coefficients are presented for characterizing the cracked sound pulses. The original high-dimensional features are adaptively optimized using principal component analysis. A hybrid framework of a support vector machine with k nearest neighbors is designed to recognize the cracked sound pulses. Finally, experiments are conducted in a practical diamond workshop to validate the feasibility and efficiency of the proposed method.