Extract Mel Frequency Cepstral Coefficients Research Articles

Baby Crying Sound Classification using Convolutional Neural Network

Crying is a crucial means of communication for newborns, crying is a newborn's early form of communication. Many individuals are unable to recognise a baby's intention from cry unless they have the appropriate training or expertise, such as nurses, paediatricians, and childcare professionals. Accurately interpreting a baby's cry can be challenging. In this research paper, the study uses a method for classifying baby crying sounds using a Convolutional Neural Network (CNN) and the dataset includes belly pain, burping, discomfort, hungry, and tired for total of 3,495 one-second-long audio clips. The research methodology involves preprocessing the audio data, extracting Mel-Frequency Cepstral Coefficients (MFCC) as features, and training the CNN model. To determine the optimal architecture, two different configurations of the CNN model are evaluated. The settings for both configurations are the same, except for the layers. The first configuration utilizes 100, 200, and 100 neurons for the respective layers, while the second configuration employs 256, 512, and 256 neurons for each layer. the results have already been evaluated that the second configuration, with deeper and more complex layers, achieves higher accuracy (86%) compared to the first configuration (84%). The study demonstrates the effectiveness of CNNs in classifying baby cries and highlights the importance of model architecture in achieving accurate classification results. Future research could explore larger and more diverse datasets to improve generalizability.

Data augmentation techniques for classification of music genre

Music genre classification, a crucial aspect in the field of audio processing, has been widely investigated using different machine learning methods. However, there is still a need to determine the most optimal application between Long Short-Term Memory (LSTM) networks and Convolutional Neural Networks (CNNs) in the field of music genre classification, and the augmentation and enhancement of the data utilized for these machine learning techniques is still an important topic to study. The objective of this research is to identify the most effective solution and to improve the accuracy of music categorization. This will be achieved through an in-depth comparison of various data augmentation techniques and neural networks. In this study, both CNN and LSTM networks are optimized and trained. The dataset has been enhanced by extracting Mel Frequency Cepstral Coefficients (MFCC) data from music recordings using three different sampling techniques. The thorough approach undertaken facilitates a detailed assessment of the effectiveness of each method in classifying genres. The experimental findings show that the combination of CNN and random sampling outperforms all other tested algorithms, resulting in a significant enhancement in genre classification accuracy. This research provides valuable findings that can inform future studies in the pursuit of more effective techniques for classifying music genres.

Analysis of Machine Learning Classifiers for Speaker Identification: A Study on SVM, Random Forest, KNN, and Decision Tree

This study investigates the performance of machine learning classifiers in the domain of speaker identification, a pivotal component of modern digital security systems. With the burgeoning integration of voice-activated interfaces in technology, the demand for accurate and reliable speaker identification is paramount. This research provides a comprehensive comparison of four widely used classifiers: Support Vector Machine (SVM), Random Forest (RF), K-Nearest Neighbors (KNN), and Decision Tree (DT). Utilizing the LibriSpeech dataset, known for its diversity of speakers and recording conditions, we extracted Mel-frequency cepstral coefficients (MFCCs) to serve as features for training and evaluating the classifiers. Each model's performance was assessed based on precision, recall, F1-score, and accuracy. The results revealed that RF outperformed all other classifiers, achieving near-perfect metrics, indicative of its robustness and generalizability for speaker identification tasks. KNN also demonstrated high performance, suggesting its suitability for applications where rapid execution and interpretability are critical. Conversely, SVM and DT, while yielding moderate and lower performances respectively, highlighted the necessity for further optimization. These findings underscore the effectiveness of ensemble and distance-based classifiers in handling complex patterns for speaker differentiation. The study not only guides the selection of appropriate classifiers for speaker identification but also sets the stage for future research, which could explore hybrid models and the impact of dataset variability on performance. The insights from this analysis contribute significantly to the field, providing a benchmark for developing advanced speaker identification systems

A Feature Selection Algorithm Based on Differential Evolution for English Speech Emotion Recognition

The automatic identification of emotions from speech holds significance in facilitating interactions between humans and machines. To improve the recognition accuracy of speech emotion, we extract mel-frequency cepstral coefficients (MFCCs) and pitch features from raw signals, and an improved differential evolution (DE) algorithm is utilized for feature selection based on K-nearest neighbor (KNN) and random forest (RF) classifiers. The proposed multivariate DE (MDE) adopts three mutation strategies to solve the slow convergence of the classical DE and maintain population diversity, and employs a jumping method to avoid falling into local traps. The simulations are conducted on four public English speech emotion datasets: eNTERFACE05, Ryerson Audio-Visual Database of Emotional Speech and Song (RAVDESS), Surrey Audio-Visual Expressed Emotion (SAEE), and Toronto Emotional Speech Set (TESS), and they cover a diverse range of emotions. The MDE algorithm is compared with PSO-assisted biogeography-based optimization (BBO_PSO), DE, and the sine cosine algorithm (SCA) on emotion recognition error, number of selected features, and running time. From the results obtained, MDE obtains the errors of 0.5270, 0.5044, 0.4490, and 0.0420 in eNTERFACE05, RAVDESS, SAVEE, and TESS based on the KNN classifier, and the errors of 0.4721, 0.4264, 0.3283 and 0.0114 based on the RF classifier. The proposed algorithm demonstrates excellent performance in emotion recognition accuracy, and it finds meaningful acoustic features from MFCCs and pitch.

Investigation of self-supervised pre-trained models for classification of voice quality from speech and neck surface accelerometer signals

Prior studies in the automatic classification of voice quality have mainly studied the use of the acoustic speech signal as input. Recently, a few studies have been carried out by jointly using both speech and neck surface accelerometer (NSA) signals as inputs, and by extracting mel-frequency cepstral coefficients (MFCCs) and glottal source features. This study examines simultaneously-recorded speech and NSA signals in the classification of voice quality (breathy, modal, and pressed) using features derived from three self-supervised pre-trained models (wav2vec2-BASE, wav2vec2-LARGE, and HuBERT) and using a support vector machine (SVM) as well as convolutional neural networks (CNNs) as classifiers. Furthermore, the effectiveness of the pre-trained models is compared in feature extraction between glottal source waveforms and raw signal waveforms for both speech and NSA inputs. Using two signal processing methods (quasi-closed phase (QCP) glottal inverse filtering and zero frequency filtering (ZFF)), glottal source waveforms are estimated from both speech and NSA signals. The study has three main goals: (1) to study whether features derived from pre-trained models improve classification accuracy compared to conventional features (spectrogram, mel-spectrogram, MFCCs, i-vector, and x-vector), (2) to investigate which of the two modalities (speech vs. NSA) is more effective as input in the classification task with pre-trained model-based features, and (3) to evaluate whether the deep learning-based CNN classifier can enhance the classification accuracy in comparison to the SVM classifier. The results revealed that the use of the NSA input showed better classification performance compared to the speech signal. Between the features, the pre-trained model-based features showed better classification accuracies, both for speech and NSA inputs compared to the conventional features. The two classifiers performed equally well for all the pre-trained model-based features for both speech and NSA signals. It was also found that the HuBERT features performed better than the wav2vec2-BASE and wav2vec2-LARGE features for both speech and NSA inputs. In particular, when compared to the conventional features, the HuBERT features showed an absolute accuracy improvement of 3%–6% for speech and NSA signals in the classification of voice quality.

Cutting Weights of Deep Learning Models for Heart Sound Classification: Introducing a Knowledge Distillation Approach.

Cardiovascular diseases (CVDs) are the number one cause of death worldwide. In recent years, intelligent auxiliary diagnosis of CVDs based on computer audition has become a popular research field, and intelligent diagnosis technology is increasingly mature. Neural networks used to monitor CVDs are becoming more complex, requiring more computing power and memory, and are difficult to deploy in wearable devices. This paper proposes a lightweight model for classifying heart sounds based on knowledge distillation, which can be deployed in wearable devices to monitor the heart sounds of wearers. The network model is designed based on Convolutional Neural Networks (CNNs). Model performance is evaluated by extracting Mel Frequency Cepstral Coefficients (MFCCs) features from the PhysioNet/CinC Challenge 2016 dataset. The experimental results show that knowledge distillation can improve a lightweight network's accuracy, and our model performs well on the test set. Especially, when the knowledge distillation temperature is 7 and the weight α is 0.1, the accuracy is 88.5 %, the recall is 83.8 %, and the specificity is 93.6 %.Clinical relevance- A lightweight model of heart sound classification based on knowledge distillation can be deployed on various hardware devices for timely monitoring and feedback of the physical condition of patients with CVDs for timely provision of medical advice. When the model is deployed on the medical instruments of the hospital, the condition of severe and hospitalised patients can be timely fed back and clinical treatment advice can be provided to the clinicians.

Information technology for gender recognition by voice

Gender recognition from voice is a challenging problem in speech processing. This task involves extracting meaningful features from speech signals and classifying them into male or female categories. In this article, was implemented a gender recognition system using Python programming. I first recorded voice samples from both male and female speakers and extracted Mel-frequency cepstral coefficients (MFCC) as features. Then trained, a Support VectorMachine (SVM) classifier was on these features and evaluated its performance using accuracy, precision, recall, and F1-score metrics. These experiments demonstrated that proposed system should achieve high accuracy on the test set and will accurately predict the gender of a speaker based on their voice. I also explored using pre-trained models to reduce the need for large amounts of training data and found that they can provide good performance while requiring less computation. This study highlights the potential of using machine learning techniques for gender recognition from voice and can be extended to other speech processing applications.

Acoustic-based automatic speech intelligibility scoring using deep neural networks

Human-generated measures of speech intelligibility are time-intensive methods for assessing the intelligibility of speech. The purpose of the present study is to automate the assessment of speech intelligibility by developing a deep neural network that estimates a standardized intelligibility score based on acoustic input. We extracted Mel-frequency cepstral coefficients from the UW/NU IEEE sentence corpus which had been manipulated with three signal-to-noise ratios (−2, 0, 2 dB). We obtained listener transcriptions from the UAW speech intelligibility dataset and calculated the Levenshtein distance between the transcriptions and the speaker's prompt. The neural network was trained to predict the Levenshtein distance given MFCC representations of sentences. We use tenfold cross-validation to verify the accuracy of the model and investigate the correlation of the model predictions with the average human responses. We also compare our model’s accuracy with the Levenshtein distance generated by transcriptions produced by the DeepSpeech ASR model. This study investigates the reliability of deep neural networks as an alternative to human-based inference in quantifying the intelligibility of speech. We discuss the advantages and disadvantages of the different approaches to assessing speech intelligibility.

Keyword identification framework for speech communication on construction sites

Worksite communication is a key to boosting teamwork and improving worker performance on the construction worksite. Communication among workers on the construction site mostly consists of speech communication. However, construction sites are typically noisy due to construction tasks like drilling and operation of heavy equipment. Meanwhile, workers on construction sites typically represent a range of different ethnic and linguistic backgrounds and have different speaking accents. This can make it difficult for the listener to understand the speaker clearly, leading to miscommunication and errors in decision making on the construction site. Technological advancements in recent years can be leveraged to mitigate this problem. In this paper, a keyword identification framework is developed for speech communication on the construction site. For this framework, 12 hours of raw audio data containing 18 crane signalman speech commands (referred to as “keywords”) are collected. The crane signalman uses specific keywords to communicate with the crane operator and guide the crane operator in the crane operations. The 2-second audio clips (this being the approximate duration of each keyword) are extracted from the raw audio dataset, and construction site noise is added. Moreover, mel-frequency cepstral coefficients are extracted from the waveform audio dataset. The extracted mel-frequency cepstral coefficients, in turn, are used to train the 1-dimensional convolutional neural network. After training, the model is found to achieve a training accuracy of 97.3%, a validation accuracy of 96.1%, and a testing accuracy of 93.8%. The model is further deployed for real-time identification of keywords in speech, with the model achieving an accuracy of 95.3%. In light of these findings, it can be concluded that the developed framework is suitable for real-time application in noisy construction sites for identifying specific keywords in speech.

A Large-Scale Benchmark Dataset for Anomaly Detection and Rare Event Classification for Audio Forensics

With the emergence of new digital technologies, a significant surge has been seen in the volume of multimedia data generated from various smart devices. Several challenges for data analysis have emerged to extract useful information from multimedia data. One such challenge is the early and accurate detection of anomalies in multimedia data. This study proposes an efficient technique for anomaly detection and classification of rare events in audio data. In this paper, we develop a vast audio dataset containing seven different rare events (anomalies) with 15 different background environmental settings (e.g., beach, restaurant, and train) to focus on both detection of anomalous audio and classification of rare sound (e.g., events—baby cry, gunshots, broken glasses, footsteps) events for audio forensics. The proposed approach uses the supreme feature extraction technique by extracting mel-frequency cepstral coefficients (MFCCs) features from the audio signals of the newly created dataset and selects the minimum number of best-performing features for optimum performance using principal component analysis (PCA). These features are input to state-of-the-art machine learning algorithms for performance analysis. We also apply machine learning algorithms to the state-of-the-art dataset and realize good results. Experimental results reveal that the proposed approach effectively detects all anomalies and superior performance to existing approaches in all environments and cases.

Design and Implementation of Intelligent Singer Recognition System

An intelligent singer recognition system was designed to identify the singer. The scheme established a song library at first, then used MATLAB to extract Mel Frequency Cepstral Coefficients (MFCC) from each song in the song library, moreover, set up characteristic parameters pattern base and trained the pattern base by Vector Quantization (VQ) to obtain the final codebook base. Finally, it can correctly classify the singer based on Dynamic Time Warping (DTW) matching reference characteristic parameters pattern with test pattern. Test results showed that the system’s recognition rate is up to 90%.

Speech emotion recognition with deep convolutional neural networks

The speech emotion recognition (or, classification) is one of the most challenging topics in data science. In this work, we introduce a new architecture, which extracts mel-frequency cepstral coefficients, chromagram, mel-scale spectrogram, Tonnetz representation, and spectral contrast features from sound files and uses them as inputs for the one-dimensional Convolutional Neural Network for the identification of emotions using samples from the Ryerson Audio-Visual Database of Emotional Speech and Song (RAVDESS), Berlin (EMO-DB), and Interactive Emotional Dyadic Motion Capture (IEMOCAP) datasets. We utilize an incremental method for modifying our initial model in order to improve classification accuracy. All of the proposed models work directly with raw sound data without the need for conversion to visual representations, unlike some previous approaches. Based on experimental results, our best-performing model outperforms existing frameworks for RAVDESS and IEMOCAP, thus setting the new state-of-the-art. For the EMO-DB dataset, it outperforms all previous works except one but compares favorably with that one in terms of generality, simplicity, and applicability. Specifically, the proposed framework obtains 71.61% for RAVDESS with 8 classes, 86.1% for EMO-DB with 535 samples in 7 classes, 95.71% for EMO-DB with 520 samples in 7 classes, and 64.3% for IEMOCAP with 4 classes in speaker-independent audio classification tasks.

A Robust Feature Extraction Method for Real-Time Speech Recognition System on a Raspberry Pi 3 Board

The development of a real-time automatic speech recognition system (ASR) better adapted to environmental variabilities, such as noisy surroundings, speaker variations and accents has become a high priority. Robustness is required, and it can be performed at the feature extraction stage which avoids the need for other pre-processing steps. In this paper, a new robust feature extraction method for real-time ASR system is presented. A combination of Mel-frequency cepstral coefficients (MFCC) and discrete wavelet transform (DWT) is proposed. This hybrid system can conserve more extracted speech features which tend to be invariant to noise. The main idea is to extract MFCC features by denoising the obtained coefficients in the wavelet domain by using a median filter (MF). The proposed system has been implemented on Raspberry Pi 3 which is a suitable platform for real-time requirements. The experiments showed a high recognition rate (100%) in clean environment and satisfying results (ranging from 80% to 100%) in noisy environments at different signal to noise ratios (SNRs).

Recognition of Spoken Languages from Acoustic Speech Signals Using Fourier Parameters

Spoken language identification (LID) or spoken language recognition (LR) is defined as the process of recognizing the language from speech utterance. In this paper, a new Fourier parameter (FP) model is proposed for the task of speaker-independent spoken language recognition. The performance of the proposed FP features is analyzed and compared with the legacy mel-frequency cepstral coefficient (MFCC) features. Two multilingual databases, namely Indian Institute of Technology Kharagpur Multilingual Indian Language Speech Corpus (IITKGP-MLILSC) and Oriental Language Recognition Speech Corpus (AP18-OLR), are used to extract FP and MFCC features. Spoken LID/LR models are developed with the extracted FP and MFCC features using three classifiers, namely support vector machines, feed-forward artificial neural networks, and deep neural networks. Experimental results show that the proposed FP features can effectively recognize different languages from speech signals. It can also be observed that the recognition performance is significantly improved when compared to MFCC features. Further, the recognition performance is enhanced when MFCC and FP features are combined.

Diagnosis of Parkinson’s Disease based on Wavelet Transform and Mel Frequency Cepstral Coefficients

The aim of this study presented in this paper is to determine the choice of the appropriate wavelet analyzer with the method of extraction of MFCC coefficients for an assistance in the diagnosis of Parkinson's disease. The analysis used is based on a database of 18 healthy and 20 Parkinsonian patients. The suggested processing is based on the transformation of the speech signal by the wavelet transform through testing several sorts of wavelets, extracting Mel Frequency Cepstral Coefficients (MFCC) from the signals, and we apply the support vector machine (SVM) as classifier. The test results reveal that the best recognition rate, which is 86.84%, is obtained by the wavelets of level 2 at 3rd scale (Daubechie, Symlet, ReverseBior or BiorSpline) combination-MFCC–SVM.

Comparative Study of MFCC Feature with Different Machine Learning Techniques in Acoustic Scene Classification

The task of labeling the audio sample in outdoor condition or indoor condition is called Acoustic Scene Classification (ASC). The ASC use acoustic information to imply about the context of the recorded environment. Since ASC can only applied in indoor environment in real world, a new set of strategies and classification techniques are required to consider for outdoor environment. In this paper, we present the comparative study of different machine learning classifiers with Mel-Frequency Cepstral Coefficients (MFCC) feature. We used DCASE Challenge 2016 dataset to show the properties of machine learning classifiers. There are several classifiers to address the ASC task. In this paper, we compare the properties of different classifiers: K-nearest neighbors (KNN), Support Vector Machine (SVM), Decision Tree (ID3) and Linear Discriminant Analysis by using MFCC feature. The best of classification methodology and feature extraction are essential for ASC task. In this comparative study, we extract MFCC feature from acoustic scene audio and then extracted feature is applied in different classifiers to know the advantages of classifiers for MFCC feature. This paper also proposed the MFCC-moment feature for ASC task by considering the statistical moment information of MFCC feature.

Detection of Pathological Voice Using Cepstrum Vectors: A Deep Learning Approach

Computerized detection of voice disorders has attracted considerable academic and clinical interest in the hope of providing an effective screening method for voice diseases before endoscopic confirmation. This study proposes a deep-learning-based approach to detect pathological voice and examines its performance and utility compared with other automatic classification algorithms. This study retrospectively collected 60 normal voice samples and 402 pathological voice samples of 8 common clinical voice disorders in a voice clinic of a tertiary teaching hospital. We extracted Mel frequency cepstral coefficients from 3-second samples of a sustained vowel. The performances of three machine learning algorithms, namely, deep neural network (DNN), support vector machine, and Gaussian mixture model, were evaluated based on a fivefold cross-validation. Collective cases from the voice disorder database of MEEI (Massachusetts Eye and Ear Infirmary) were used to verify the performance of the classification mechanisms. The experimental results demonstrated that DNN outperforms Gaussian mixture model and support vector machine. Its accuracy in detecting voice pathologies reached 94.26% and 90.52% in male and female subjects, based on three representative Mel frequency cepstral coefficient features. When applied to the MEEI database for validation, the DNN also achieved a higher accuracy (99.32%) than the other two classification algorithms. By stacking several layers of neurons with optimized weights, the proposed DNN algorithm can fully utilize the acoustic features and efficiently differentiate between normal and pathological voice samples. Based on this pilot study, future research may proceed to explore more application of DNN from laboratory and clinical perspectives.

Audio-cough event detection based on moment theory

Cough detection has recently been identified as of paramount importance to fully exploit the potential of telemedicine in respiratory conditions and release some of the economic burden of respiratory care in national health systems. Current audio-based cough detection systems are either uncomfortable or not suitable for continuous patient monitoring, since the audio processing methods implemented therein fail to cope with noisy environments such as those where the acquiring device is carried in the pocket (e.g. smartphone). Moment theory has been widely applied in a number of complex problems involving image processing, computer vision, and pattern recognition. Their invariance properties and noise robustness make them especially suitable as “signature” features enabling character recognition or texture analysis. A natural extension of moment theory to one-dimensional signals is the identification of meaningful patterns in audio signals. However, to the best of our knowledge only marginal attempts have been made in this direction. This paper applies moment theory to perform cough detection in noisy audio signals. Our proposal adopts the first steps used to extract Mel frequency cepstral coefficients (time-frequency decomposition and application of a filter bank defined in the Mel scale) while the innovation is introduced in the second step, where energy patterns defined for specific temporal frames and frequency bands are characterised using moment theory. Our results show the feasibility of using moment theory to solve such a problem in a variety of noise conditions, with sensitivity and specificity values around 90%, significantly outperforming popular state-of-the-art feature sets.

Improved Detection of Lung Fluid With Standardized Acoustic Stimulation of the Chest.

Accumulation of excess air and water in the lungs leads to breakdown of respiratory function and is a common cause of patient hospitalization. Compact and non-invasive methods to detect the changes in lung fluid accumulation can allow physicians to assess patients’ respiratory conditions. In this paper, an acoustic transducer and a digital stethoscope system are proposed as a targeted solution for this clinical need. Alterations in the structure of the lungs lead to measurable changes which can be used to assess lung pathology. We standardize this procedure by sending a controlled signal through the lungs of six healthy subjects and six patients with lung disease. We extract mel-frequency cepstral coefficients and spectroid audio features, commonly used in classification for music retrieval, to characterize subjects as healthy or diseased. Using the n}{}Kn-nearest neighbors algorithm, we demonstrate 91.7% accuracy in distinguishing between healthy subjects and patients with lung pathology.

Comparison of Feature Extraction Mel Frequency Cepstral Coefficients and Linear Predictive Coding in Automatic Speech Recognition for Indonesian

Speech recognition can be defined as the process of converting voice signals into the ranks of the word, by applying a specific algorithm that is implemented in a computer program. The research of speech recognition in Indonesia is relatively limited. This paper has studied methods of feature extraction which is the best among the Linear Predictive Coding (LPC) and Mel Frequency Cepstral Coefficients (MFCC) for speech recognition in Indonesian language. This is important because the method can produce a high accuracy for a particular language does not necessarily produce the same accuracy for other languages, considering every language has different characteristics. Thus this research hopefully can help further accelerate the use of automatic speech recognition for Indonesian language. There are two main processes in speech recognition, feature extraction and recognition. The method used for comparison feature extraction in this study is the LPC and MFCC, while the method of recognition using Hidden Markov Model (HMM). The test results showed that the LPC method is better than MFCC in Indonesian language speech recognition.