Mel Frequency Cepstrum Coefficient Feature Research Articles

Speech emotion recognition using multiple classification models based on MFCC feature values

In everyday life, people are often influenced by emotions in our behaviour and language. When people use different emotions to articulate the same text, it can have a completely different effect. With the increasing demand for speech emotion recognition (SER), more machine learning and deep learning methods are being used to perform SER. matlab was used as the experimental tool in this study. The Berlin Database of Emotional Speech was used as the database. Feature extraction was performed by Mel Frequency Cepstrum Coefficients (MFCC) and based on these feature values, Support Vector Machines (SVM), K-Nearest Neighbors Algorithm (KNN), Semi-supervised graph-based classifier, ECOC classification model, Naive Bayes model and long short-term memory for predictive classification. The results surface that among the six classifiers, the best sentiment recognition method is LSTM, with 93.2% and 73.03% accuracies in the training and test groups respectively.

Coal-gangue recognition via multi-branch convolutional neural network based on MFCC in noisy environment

Traditional coal-gangue recognition methods usually do not consider the impact of equipment noise, which severely limits its adaptability and recognition accuracy. This paper mainly studies the more accurate recognition of coal-gangue in the noise site environment with the operation of shearer, conveyor, transfer machine and other device in the process of top coal caving. Mel Frequency Cepstrum Coefficients (MFCC) smoothing method was introduced to express the intrinsic feature of sound pressure more clearly in the coal-gangue recognition site. Then, a multi-branch convolution neural network (MBCNN) model with three branches was developed, and the smoothed MFCC feature was incorporated into this model to realize the recognition of falling coal and gangue in noisy environment. The sound pressure signal datasets under the operation of different device were constructed through a great deal of laboratory and site data acquisition. Comparative experiments were carried out on noiseless dataset, single noise dataset and simulated site dataset, and the results show that our method can provide higher correct recognition accuracy and better robustness. The proposed coal-gangue recognition approach based on MBCNN and MFCC smoothing can not only recognize the state of falling coal or gangue, but also recognize the operational state of site device.

Research on the mechanical fault diagnosis method based on sound signal and IEMD-DDCNN

PurposeThe purpose of this paper is to provide a shorter time cost, high-accuracy fault diagnosis method for water pumps. Water pumps are widely used in industrial equipment and their fault diagnosis is gaining increasing attention. Considering the time-consuming empirical mode decomposition (EMD) method and the more efficient classification provided by the convolutional neural network (CNN) method, a novel classification method based on incomplete empirical mode decomposition (IEMD) and dual-input dual-channel convolutional neural network (DDCNN) composite data is proposed and applied to the fault diagnosis of water pumps.Design/methodology/approachThis paper proposes a data preprocessing method using IEMD combined with mel-frequency cepstrum coefficient (MFCC) and a neural network model of DDCNN. First, the sound signal is decomposed by IEMD to get numerous intrinsic mode functions (IMFs) and a residual (RES). Several IMFs and one RES are then extracted by MFCC features. Ultimately, the obtained features are split into two channels (IMFs one channel; RES one channel) and input into DDCNN.FindingsThe Sound Dataset for Malfunctioning Industrial Machine Investigation and Inspection (MIMII dataset) is used to verify the practicability of the method. Experimental results show that decomposition into an IMF is optimal when taking into account the real-time and accuracy of the diagnosis. Compared with EMD, 51.52% of data preprocessing time, 67.25% of network training time and 63.7% of test time are saved and also improve accuracy.Research limitations/implicationsThis method can achieve higher accuracy in fault diagnosis with a shorter time cost. Therefore, the fault diagnosis of equipment based on the sound signal in the factory has certain feasibility and research importance.Originality/valueThis method provides a feasible method for mechanical fault diagnosis based on sound signals in industrial applications.

Faulty voice diagnosis of automotive gearbox based on acoustic feature extraction and classification technique

This research work objectives to establish a systematic approach to automobile gearbox defect diagnostics by acquiring the acoustic feature from their faulty sound information using efficient machine learning techniques. Machine learning method is important for automated fault diagnosis because it uses acquired data to identify machine conditions and defects from their sound. This motivates studying machine learning for gearbox defect diagnosis using acoustic analysis, and their sound inputs. Unexpected gear and bearing failures result in considerable system damage, significant financial loss, and significant safety concerns, in addition to significant system damage. The constant operation and fluctuating cyclic loading conditions can result in catastrophic failure of the gearbox. Utilising condition monitoring, predictive maintenance of a gearbox has been practised over time. Numerous methods, including acoustic analysis from their normal sound and some faulty sound in gear box and others, can be used to conduct non-destructive testing. Following the creation of an experimental setup for several classes of gearbox defects, acoustic features are extracted from the gear in the form of statistical coefficients in term of mean and correlation coefficients. The acoustic features are calculated using Mel-frequency cepstrum coefficients (MFCC) and their derivatives ∆MFCC and ∆∆ MFCC feature extraction technique. The best statistical features are chosen using support vector machine (SVM), decision tree (DT), linear decrement analysis (LDA), Naive Bayes (NB) and Logarithm regression (LR) classifiers. All of the mentioned classifiers here SVM and NB are given the best classification result. From the features based classifiers techniques, the defective component was found.

TFA-CLSTMNN: Novel convolutional network for sound-based diagnosis of COVID-19

The outbreak of the global COVID-19 pandemic has become a public crisis and is threatening human life in every country. Recently, researchers have developed testing methods via patients cough recordings. In order to improve the testing accuracy, in this paper, we establish a novel COVID-19 sound-based diagnosis framework, i.e. TFA-CLSTMNN, which integrates time-frequency domain features of the recorded cough with the Attention-Convolution Long Short-Term Memory Neural Network. Specifically, we calculate the Mel-frequency cepstrum coefficient (MFCC) of the cough data to extract the time-frequency domain features. We then apply the convolutional neural network and the attentional mechanism on the time-frequency features, which is followed by the long short-term memory neural network to analyze the MFCC features of the data. The recognition and classification can be then carried out to evaluate the positiveness or negativeness of the tested samples. Experimental results show that the proposed TFA-CLSTMNN framework outperforms the baseline neural networks in sound-based COVID-19 diagnosis and derives an accuracy over 0.95 on the public real-world datasets.

Design and Implementation of Energy-Efficient Floating Point MFCC Extraction Architecture for Speech Recognition Systems

Abstract: This brief presents an energy-efficient architecture to extract mel-frequency cepstrum coefficients (MFCCs) for realtime speech recognition systems. Based on the algorithmic property of MFCC feature extraction, the architecture is designed with floating-point arithmetic units to cover a wide dynamic range with a small bit-width. Moreover, various operations required in the MFCC extraction are examined to optimize operational bit-width and lookup tables needed to compute nonlinear functions, such as trigonometric and logarithmic functions. In addition, the dataflow of MFCC extraction is tailored to minimize the computation time. As a result, the energy consumption is considerably reduced compared with previous MFCC extraction systems

Heart Sound Classification Based on Mel-Frequency Cepstrum Coefficient Features and Multi-Scale Residual Recurrent Neural Networks

A rapid and accurate algorithm model of extracting heart sounds plays a vital role in the early detection of cardiovascular disorders, especially for small primary health care clinics. This paper proposes a heart sound extraction and classification algorithm based on static and dynamic combination of Mel-frequency cepstrum coefficient (MFCC) feature extraction and the multi-scale residual recurrent neural network (MsRes-RNN) algorithm model. The standard MFCC parameters represent the static characteristics of the signal. In contrast, the first-order and second-order MFCC parameters represent the dynamic characteristics of the signal. They are extracted and combined to form the MFCC feature representation. Then, the MFCC-based features are fed to a MsRes-RNN algorithm model for feature learning and classification tasks. The proposed classification model can take advantage of the encoded local characteristics extracted from the multi-scale residual neural network (MsResNet) and the long-term dependencies captured by recurrent neural network (RNN). Model estimation experiments and performance comparisons with other state-of-the-art algorithms are presented in this paper. Experiments indicate that a classification accuracy of 93.9% has been achieved on 2016 PhysioNet/CinC Challenge datasets.

Metaverse Space Ecological Scene Design Based on Multimedia Digital Technology

Yuan cosmos is a virtual world linked and created by scientific and technological means, which is mapped and interacted with the real world, and a digital living space with a new social system. With the increasing popularity of data acquisition and production equipment, people are increasingly convenient to produce multimedia data such as images, graphics, audio, video, animation, and three-dimensional models. In addition to the rapid development of digital technology itself, the biological information technology related to digital technology also greatly promotes the emergence of the metauniverse. This paper aims to study the application of multimedia digital technology to the ecological scene design of metauniverse space, introduces the related concepts of metauniverse and multimedia digital technology, expounds the related methods of multimedia digital technology and neural network related algorithms, and then takes the three-dimensional simulation of the auditory system in the interactive multisensory simulation system of the constituent elements of metauniverse as an example. The mel-frequency cepstrum coefficient (MFCC) is used to simulate the auditory characteristics of the auditory periphery (cochlea) as the perceptual end of the model. A variety of bionic mechanisms are used in the model, such as designing the connection mode of neurons, learning state and release effect, and the regeneration mechanism of neurons. For the verification of the performance of the model, the speech sample database, including English words and phrases, is recorded and the speech content information recognized by the model by means of speech recognition is experimented. The experimental results show that, in terms of phrase accuracy, the DN-1 model improves 2.59% and the DN-2 model improves 2.77% compared with MFCC feature on the basis of mixed features. When only DBN features are used, the performance improvement rate of the developmental network model is small.

Time-Delay-Neural-Network-Based Audio Feature Extractor for Ultra-Low Power Keyword Spotting

In this brief, we propose an audio feature extractor, based on a time delay neural network (TDNN), for ultra-low power keyword spotting (KWS). Conventionally, mel-frequency cepstrum coefficients (MFCCs) are widely used as features for KWS. However, an analog-to-digital converter (ADC) with high precision and high sampling rate is required for computing MFCCs, which consumes large amount of power. In our proposed feature extractor, on the other hand, input audio signals are band-pass filtered, and then the filtered signals are processed in TDNN in the analog domain, which can substitute a high precision ADC with simple clocked comparators. Simulation results show that the power dissipation of the proposed feature extractor can be reduced by 88% compared to the conventional MFCC feature extractor.

Efficiently Classifying Lung Sounds through Depthwise Separable CNN Models with Fused STFT and MFCC Features.

Lung sounds remain vital in clinical diagnosis as they reveal associations with pulmonary pathologies. With COVID-19 spreading across the world, it has become more pressing for medical professionals to better leverage artificial intelligence for faster and more accurate lung auscultation. This research aims to propose a feature engineering process that extracts the dedicated features for the depthwise separable convolution neural network (DS-CNN) to classify lung sounds accurately and efficiently. We extracted a total of three features for the shrunk DS-CNN model: the short-time Fourier-transformed (STFT) feature, the Mel-frequency cepstrum coefficient (MFCC) feature, and the fused features of these two. We observed that while DS-CNN models trained on either the STFT or the MFCC feature achieved an accuracy of 82.27% and 73.02%, respectively, fusing both features led to a higher accuracy of 85.74%. In addition, our method achieved 16 times higher inference speed on an edge device and only 0.45% less accuracy than RespireNet. This finding indicates that the fusion of the STFT and MFCC features and DS-CNN would be a model design for lightweight edge devices to achieve accurate AI-aided detection of lung diseases.

Thunder Signal Detection via Deep Learning

In this paper, a thunder signal detection method is proposed based on the deep learning framework. The recorded thunder signal is segment-wise acquired, stored and pre-processed. In each frame, we use Mel Frequency Cepstrum Coefficient (MFCC) to extract the features of the thunder signal, which is consistent with the frequency characteristics of human perception. We then use the MFCC features derived in each frame to form a 3-channel tensor data, which is used as the further input to the designed convolutional neural network (CNN). The goal of CNN is to classify the existence of thunder for a single data frame. To improve the robustness of CNN, we included other confusing signals that are similar to thunder signals in the training and testing datasets. On the testing dataset, our proposed method outperforms the state-of-art methods in terms of accuracy, sensitivity, and specificity. Our proposed deep-learning-based thunder detection method not only increases the real-time performance of the lighting location system with thunder signals but also further improves the accuracy of other sound alarm systems.

Windturbine blade fault diagnosis based on MFCC feature optimization strategy

The fault problem of windturbine blades cannot be ignored. This paper proposes a fault diagnosis method for windturbine blades based on feature optimization strategy. First, the Mel-frequency cepstrum coefficient(MFCC) method is used to extract the feature samples of the windturbine sound signal, and then the entropy weight method is used to optimize the characteristics of the feature samples, and finally the K-means clustering is combined to realize the fault diagnosis of the windturbine blade. The effectiveness of the proposed method is verified by using two faulty windturbine and one faultless windturbine in a wind farm.

Identification of the English Accent Spoken in Different Countries by the k-Nearest Neighbor Method

Sound is the pressure wave created by an object vibrating with a certain frequency. 3 organs are needed for the formation of voice in humans. These are lungs, vocal cords and mouth. Due to the structure of these organs and the similarity of the person with their current language, they can speak another language with different accent. A language can be spoken in different parts of the same country and in different countries. The second most widely used language in the world is English, has numerous accents around the world. In this study, it is aimed to determine which country the English accent spoken in different regions belongs to. In the dataset used, there are 330 sound samples including English accents spoken in Spain, France, Germany, Italy, England and America. Classification has been made with 12 features obtained by Mel Frequency Cepstrum Coefficients feature extraction method. k-Nearest Neighbor (kNN) were used in the classification and 87.2% success was achieved.

Emotion recognition using speech and neural structured learning to facilitate edge intelligence

Emotions are quite important in our daily communications and recent years have witnessed a lot of research works to develop reliable emotion recognition systems based on various types data sources such as audio and video. Since there is no apparently visual information of human faces, emotion analysis based on only audio data is a very challenging task. In this work, a novel emotion recognition is proposed based on robust features and machine learning from audio speech. For a person independent emotion recognition system, audio data is used as input to the system from which, Mel Frequency Cepstrum Coefficients (MFCC) are calculated as features. The MFCC features are then followed by discriminant analysis to minimize the inner-class scatterings while maximizing the inter-class scatterings. The robust discriminant features are then applied with an efficient and fast deep learning approach Neural Structured Learning (NSL) for emotion training and recognition. The proposed approach of combining MFCC, discriminant analysis and NSL generated superior recognition rates compared to other traditional approaches such as MFCC-DBN, MFCC-CNN, and MFCC-RNN during the experiments on an emotion dataset of audio speeches. The system can be adopted in smart environments such as homes or clinics to provide affective healthcare. Since NSL is fast and easy to implement, it can be tried on edge devices with limited datasets collected from edge sensors. Hence, we can push the decision-making step towards where data resides rather than conventionally processing of data and making decisions from far away of the data sources. The proposed approach can be applied in different practical applications such as understanding peoples’ emotions in their daily life and stress from the voice of the pilots or air traffic controllers in air traffic management systems.

Heart sound classification based on improved MFCC features and convolutional recurrent neural networks

Heart sound classification plays a vital role in the early detection of cardiovascular disorders, especially for small primary health care clinics. Despite that much progress has been made for heart sound classification in recent years, most of them are based on conventional segmented features and shallow structure based classifiers. These conventional acoustic representation and classification methods may be insufficient in characterizing heart sound, and generally suffer from a degraded performance due to the complicated and changeable cardiac acoustic environment. In this paper, we propose a new heart sound classification method based on improved Mel-frequency cepstrum coefficient (MFCC) features and convolutional recurrent neural networks. The Mel-frequency cepstrums are firstly calculated without dividing the heart sound signal. A new improved feature extraction scheme based on MFCC is proposed to elaborate the dynamic characteristics among consecutive heart sound signals. Finally, the MFCC-based features are fed to a deep convolutional and recurrent neural network (CRNN) for feature learning and later classification task. The proposed deep learning framework can take advantage of the encoded local characteristics extracted from the convolutional neural network (CNN) and the long-term dependencies captured by the recurrent neural network (RNN). Comprehensive studies on the performance of different network parameters and different network connection strategies are presented in this paper. Performance comparisons with state-of-the-art algorithms are given for discussions. Experiments show that, for the two-class classification problem (pathological or non-pathological), a classification accuracy of 98% has been achieved on the 2016 PhysioNet/CinC Challenge database.

Personalized Power Consumption Control Method for Intelligent Buildings That Perceiving Speech Recognition Identity

With the development of IoT technology, the meaning of the concept of energy conservation and environmental protection in people’s minds is also constantly changing, and it is moving towards the direction of intelligence and personalization. At present, China’s economy has entered the era of artificial intelligence. Electricity control such as building lighting has gradually changed from a single manual control to a multi-modal control. This paper designs an embedded control method based on speech recognition of personnel to achieve personalized power consumption, and improves the existing defects of intelligent power consumption in buildings such as unconfirmed operators and single decision-making factors.The method is to extract the mel frequency cepstrum coefficient feature vector after preprocessing the speech signal; Aiming at the problem of singularity matrix in the parameter estimation of Gaussian mixture model, the modified EM algorithm was used to optimize the parameter estimation of the Gaussian mixture model to complete the identification of personnel. The experimental test results show that the system can control the electricity facilities in combination with the environmental information and realize the personalized lighting control function while confirming the identity of the personnel.

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.

Multimedia utilization of non-computerized disguised voice and acoustic similarity measurement

In this paper, an identification of a speaker for multimedia application under non-electronically disguised voice is performed. In non-electronically disguised voice under physical variation of speech, it is a difficult task to identify the speaker in speech signal processing application area. Due to changes in the frequency spectrum of the speech signal during non- electronic disguising, some methods like Mel-frequency cepstrum coefficients (MFCC), delta Mel-frequency cepstrum coefficients (ΔMFCC) and double delta Mel-frequency cepstrum coefficients (ΔΔMFCC) are used to specify the frequencies spectral property. A new algorithm developed, based on acoustic feature extraction by MFCC technique of text-dependent speech signal of all speaker’s and changed their speech by six physical variation methods. The acoustic features which include the correlation coefficients and the mean value are extracted by the MFCC, ΔMFCC and ΔΔMFCC feature extraction method. Thereafter, different classifiers based on feature extraction are used to classify the non-electronically disguised voice and normal voice.

Robustness Speaker Recognition Based on Feature Space in Clean and Noisy Condition

Background & Objective: Speaker Recognition (SR) techniques have been developed into a relatively mature status over the past few decades through development work. Existing methods typically use robust features extracted from clean speech signals, and therefore in idealized conditions can achieve very high recognition accuracy. For critical applications, such as security and forensics, robustness and reliability of the system are crucial. Methods: The background noise and reverberation as often occur in many real-world applications are known to compromise recognition performance. To improve the performance of speaker verification systems, an effective and robust technique is proposed to extract features for speech processing, capable of operating in the clean and noisy condition. Mel Frequency Cepstrum Coefficients (MFCCs) and Gammatone Frequency Cepstral Coefficients (GFCC) are the mature techniques and the most common features, which are used for speaker recognition. MFCCs are calculated from the log energies in frequency bands distributed over a mel scale. While GFCC has been acquired from a bank of Gammatone filters, which was originally suggested to model human cochlear filtering. This paper investigates the performance of GFCC and the conventional MFCC feature in clean and noisy conditions. The effects of the Signal-to-Noise Ratio (SNR) and language mismatch on the system performance have been taken into account in this work. Conclusion: Experimental results have shown significant improvement in system performance in terms of reduced equal error rate and detection error trade-off. Performance in terms of recognition rates under various types of noise, various Signal-to-Noise Ratios (SNRs) was quantified via simulation. Results of the study are also presented and discussed.

An efficient heart murmur recognition and cardiovascular disorders classification system.

The problem addressed in this work is the detection of a heart murmur and the classification of the associated cardiovascular disorder based on the heart sound signal. For this purpose, a dataset of Phonocardiogram (PCG) signals is acquired using baseline conditions. The dataset is acquired from 283 volunteers using Littman 3200 electronic stethoscope for a normal and four different types of heart murmurs. The samples are labelled and validated through echocardiography test of each participating volunteer. For feature extraction, normalized average Shannon energy with time-domain characteristics of heart sound signal is exploited to segment the PCG signal into its components. To improve the quality of the features, in contrast to the previous methods, all systole and diastole intervals are utilized to extract 50 Mel-Frequency Cepstrum Coefficients (MFCC) based features. Then, the iterative backward elimination method is used to identify and remove the redundant features to reduce the complexity in order to conceive a computationally tractable system. An MFCC feature vector of dimension 26 is selected for training seven different types of Support Vector Machine (SVM) and K-Nearest Neighbors (KNN) based classifiers for detection and classification of cardiovascular disorders. Fivefold cross-validation and 20% data holdout validation schemes are used for testing the classifiers. Classification accuracy of 92.6% is achieved using selected features and medium Gaussian SVM classifier. The learning curves show a good bias-variance trade-off indicating a well-fitted and generalized model for making future predictions.