Figure 1- - uploaded by Hadjaidji Elmoundher
Content may be subject to copyright.
Source publication
Spasmodic Dysphonia (SD) is a neurological problem that involves the laryngeal muscles to malfunction. It is characterized by inappropriate contraction of the laryngeal muscles during speech. To distinguish healthy and pathological human voices, we used a variety of machine learning classifiers to conduct a side-by-side comparison for the detection...
Context in source publication
Similar publications
![The structure of human speech production [45]](publication/380151317/figure/fig1/AS:11431281239353580@1714270013506/The-structure-of-human-speech-production-45_Q320.jpg)
Voice Disorder or Dysphonia has caught the attention of audio signal process engineers and researchers. The efficiency of several feature extraction and classifier implementation techniques in identifying voice abnormalities has been investigated. Mel-Frequency Cepstral Coefficient (MFCC) has been extensively used as a feature extractor. This paper...
Citations
... As a result, researchers are more favourable to implementing SVM as a classifier in classical pipeline machine learning. Nonetheless, in [31], DT gives the best results in classifying pathological with an accuracy of 86.66% compared to SVM at 80.00% and KNN at 81.60% with the use of the SVD database of vowel 'a' with low-tone for female voice. ...
... Based on Table 2, most of the researchers considered SVM as their classifier [11,24,31,52,55]. However, based on the accuracy results, we concluded that DT is more reliable compared to SVM. ...
... However, based on the accuracy results, we concluded that DT is more reliable compared to SVM. As in [31], DT has higher accuracy than SVM, which is 86.66% and 80.00% respectively. Furthermore, SVM could achieve higher accuracy when enhanced with additional features, as shown in [11] (energy profile, MFCC derivatives) and [15] (Fundamental Frequency, Jitter, Shimmer, HNR). ...
Voice Disorder or Dysphonia has caught the attention of audio signal process engineers and researchers. The efficiency of several feature extraction and classifier implementation techniques in identifying voice abnormalities has been investigated. Mel-Frequency Cepstral Coefficient (MFCC) has been extensively used as a feature extractor. This paper adopts a Comparative Review Method to assess the effectiveness of feature extraction and classifier methods in detecting voice disorders. By examining the pairing of the Mel-Frequency Cepstral Coefficient (MFCC) with various classifiers, including Support Vector Machine (SVM), Artificial Neural Network (ANN), Decision Tree (DT), and other online or commercial classifiers, the study aims to review the robustness of MFCC in this context. The study also recognizes the significance of choosing the right database in light of the various aetiologies of pathological illnesses and its possible influence on the efficacy of voice disorder detection.
... Both SVM and KNN achieved 100% accuracy, while Levenberg BPN achieved 96.7%. Another study by [3] compared DT, SVM, and KNN for classifying sustained /a/ phonation at low pitch using 11 features. The DT algorithm achieved the best performance with 86.66% accuracy. ...
... Various research has been conducted using extracted features from speech samples to train various machine learning algorithms such as Support Vector Machine (SVM), K-Nearest Neighbour (KNN), Random Forest, Decision Tree and many others to distinguish between healthy and dysphonic speech (Hadjaidji et al., 2021;Syed et al., 2021;Wu et al., 2017). ...
In this study, cross-lingual binary classification and severity estimation of dysphonic speech have been carried out. Hand-crafted acoustic feature extraction is replaced by the speaker embedding techniques used in the speaker verification. Two state of art deep learning methods for speaker verification have been used: the X-vector and ECAPA-TDNN. Embeddings are extracted from speech samples in Hungarian and Dutch languages and used to train Support Vector Machine (SVM) and Support Vector Regressor (SVR) for binary classification and severity estimation, in a cross-language manner. Our results were competitive with manual feature engineering, when the models were trained on Hungarian samples and evaluated on Dutch samples in the binary classification of dysphonic speech and outperformed in estimating the severity level of dysphonic speech. Moreover, our model achieved 0.769 and 0.771 in Spearman and Pearson correlations. Also, our results in both classification and regression were superior compared to manual feature extraction technique when models were trained on Dutch samples and evaluated on Hungarian samples with only a limited number of samples are available for training. An accuracy of 86.8% was reached with features extracted from embedding methods, while the maximum accuracy using hand-crafted acoustic features was 66.8%. Overall results show that Emphasized Channel Attention, Propagation and Aggregation in Time Delay Neural Network (ECAPA-TDNN) performs better than the former X-vector in both tasks.
... In both cases, the input is the k training samples from the dataset closest to each other. KNN is mainly used for classification or regression [13]. ...
... Concept of SVM[13] ...
The spread of the coronavirus has claimed the lives of millions around the world, which led to the emergence of an economic and health crisis at the global level, which prompted many researchers to submit proposals for early diagnosis of the coronavirus to limit its spread. In this work, we propose an automated system to detect COVID-19 based on the cough as one of the most important infection indicators. Several studies have shown that coughing amounts to 65% of the total symptoms of infection. The proposed system is mainly based on three main steps: first, cough signal detection and segmentation. Second, cough signal extraction; and third, three techniques of supervised machine learning-based classification: Support Vector Machine (SVM), K - Nearest Neighbours (KNN), and Decision Tree (DT). Our proposed system showed high performance through good accuracy values, where the best accuracy for classifying female cough was 99.6% using KNN and 88% for males using SVM
