Voice Source Signal Research Articles

Speech enhancement system design is based on FastICA blind source separation algorithm

Based on the analysis of existing speech enhancement based on the drawbacks of the traditional algorithm is adopted in the system is proposed based on FastICA blind source separation algorithm design of speech enhancement algorithm, and the transplanted to embedded speech enhancement system. The system real-time speech enhancement, by four element microphone arrays is used to sample the space of sound signal and through the built-in speech enhancement algorithm to the voice source signal and noise source signal separation, capable of suppressing co channel noise, active noise and residual background noise. Based on the test results of the algorithm on PC platform, the system meets the design requirements and the effect is good.

The automatic detection of heart failure using speech signals

Heart failure (HF) is a major global health concern and is increasing in prevalence. It affects the larynx and breathing – thereby the quality of speech. In this article, we propose an approach for the automatic detection of people with HF using the speech signal. The proposed method explores mel-frequency cepstral coefficient (MFCC) features, glottal features, and their combination to distinguish HF from healthy speech. The glottal features were extracted from the voice source signal estimated using glottal inverse filtering. Four machine learning algorithms, namely, support vector machine, Extra Tree, AdaBoost, and feed-forward neural network (FFNN), were trained separately for individual features and their combination. It was observed that the MFCC features yielded higher classification accuracies compared to glottal features. Furthermore, the complementary nature of glottal features was investigated by combining these features with the MFCC features. Our results show that the FFNN classifier trained using a reduced set of glottal + MFCC features achieved the best overall performance in both speaker-dependent and speaker-independent scenarios.

Detection of Specific Language Impairment in Children Using Glottal Source Features

Developmental dysphasia, also known as specific language impairment (SLI), is a language disorder in children that involves difficulty in speaking and understanding spoken words. Detecting SLI at an early stage is very important for successful speech therapy in children. In this paper, we propose a novel approach based on glottal source features for detecting children with SLI using the speech signal. The proposed method utilizes time- and frequency-domain glottal parameters, which are extracted from the voice source signal obtained using glottal inverse filtering (GIF). In addition, Mel-frequency cepstral coefficient (MFCC) and openSMILE based acoustic features are also extracted from speech utterances. Two machine learning algorithms, namely, support vector machine (SVM) and feed-forward neural network (FFNN), are trained separately for the MFCC, openSMILE and glottal features. A leave-fourteen-speakers-out cross-validation strategy is used for evaluating the classifiers. The experiments are conducted using the SLI speech corpus launched by the LANNA research group. Experimental results show that the glottal parameters contain significant discriminative information required for identifying children with SLI. Furthermore, the complementary nature of glottal parameters is investigated by independently combining these features with the MFCC and openSMILE acoustic features. The overall results indicate that the glottal features when used in combination with MFCC feature set provides the best performance with the FFNN classifier in the speaker-independent scenario.

Adversarial Approximate Inference for Speech to Electroglottograph Conversion

Speech produced by human vocal apparatus conveys substantial non-semantic information including the gender of the speaker, voice quality, affective state, abnormalities in the vocal apparatus etc. Such information is attributed to the properties of the voice source signal, which is usually estimated from the speech signal. However, most of the source estimation techniques depend heavily on the goodness of the model assumptions and are prone to noise. A popular alternative is to indirectly obtain the source information through the Electroglottographic (EGG) signal that measures the electrical admittance around the vocal folds using dedicated hardware. In this paper, we address the problem of estimating the EGG signal directly from the speech signal, devoid of any hardware. Sampling from the intractable conditional distribution of the EGG signal given the speech signal is accomplished through optimization of an evidence lower bound. This is constructed via minimization of the KL-divergence between the true and the approximated posteriors of a latent variable learned using a deep neural auto-encoder that serves an informative prior. We demonstrate the efficacy of the method at generating the EGG signal by conducting several experiments on datasets comprising multiple speakers, voice qualities, noise settings and speech pathologies. The proposed method is evaluated on many benchmark metrics and is found to agree with the gold standard while proving better than the state-of-the-art algorithms on a few tasks such as epoch extraction.

Contribution of Data Augmentation for the Prenventive Detection of Vocal Fold Precancerous Lesions

Purpose: In this paper, an automatic detection of premalignant lesions based on human voice production theory is designed. The current framework is interested in particularly to the case of vocal fold precancerous lesion detection. Since the earlier detection of cancer is an important fact and directly related to medical treatment, the current paper presents a non-invasive and low time consuming technique for earlier cancer detection. Method: By a simple microphone, a speech signal can be picked up and analysed. We aim to extract the voice source signal from the acoustic speech signal. The voice source generated by vocal fold is altered when a premalignant lesion occurs. Features extracted from source voices are deeply analysed. However, due to the lack of speech samples, extracted features are augmented based on features analysis. Results: The adopted method based on boxplot, histogram and probability density leads to data augmentation of the extracted features. Augmented features are used in learning and testing, processing using SVM. The performances are assessed using four criteria, sensitivity, specificity, precision and accuracy. When augmented features are combined according to PCA analysis, an accuracy of premalignant lesion identification about 95% is accomplished. Conclusion: It is shown in this study that is possible to detect the premalignant lesions with acceptable and fairly sensitivity, specificity, precision and accuracy. The performances are improved when data augmentation process is used.

Voice Source Variation Between Vowels in Male Opera Singers

The theory of nonlinear source-filter interaction predicts that the glottal voice source should be affected by the frequency relationship between formants and partials. An attempt to experimentally verify this theory is presented. Glottal voice source and electrolaryngograph (ELG) signal differences between vowels were analyzed in vowel sequences, sung at four pitches with the same degree of vocal loudness by professional opera singers. In addition, the relationships between such differences and the frequency distance between the first formant (F1) and its closest partial were examined. A digital laryngograph microprocessor was used to simultaneously record audio and ELG signals. The former was inverse filtered, and voice source parameters and formant frequencies were extracted. The amplitude quotient of the derivative of the ELG signal (AQdELG) and the contact quotient were also compared. A one-way repeated-measures ANOVA revealed significant differences between vowels, for contact quotient at four pitches and for maximum flow declination rate (MFDR) at three pitches. For other voice source parameters, differences were found at one or two pitches only. No consistent correlation was found between MFDR and the distance between F1 and its closest partial. The glottal voice source tends to vary between vowels, presumably because of source-filter interaction, but the variation does not seem to be dependent on the frequency distance between F1 and its closest partial.

Voice source characterization using pitch synchronous discrete cosine transform for speaker identification.

A characterization of the voice source (VS) signal by the pitch synchronous (PS) discrete cosine transform (DCT) is proposed. With the integrated linear prediction residual (ILPR) as the VS estimate, the PS DCT of the ILPR is evaluated as a feature vector for speaker identification (SID). On TIMIT and YOHO databases, using a Gaussian mixture model (GMM)-based classifier, it performs on par with existing VS-based features. On the NIST 2003 database, fusion with a GMM-based classifier using MFCC features improves the identification accuracy by 12% in absolute terms, proving that the proposed characterization has good promise as a feature for SID studies.

Epoch Extraction Based on Integrated Linear Prediction Residual Using Plosion Index

Epoch is defined as the instant of significant excitation within a pitch period of voiced speech. Epoch extraction continues to attract the interest of researchers because of its significance in speech analysis. Existing high performance epoch extraction algorithms require either dynamic programming techniques or a priori information of the average pitch period. An algorithm without such requirements is proposed based on integrated linear prediction residual (ILPR) which resembles the voice source signal. Half wave rectified and negated ILPR (or Hilbert transform of ILPR) is used as the pre-processed signal. A new non-linear temporal measure named the plosion index (PI) has been proposed for detecting `transients' in speech signal. An extension of PI, called the dynamic plosion index (DPI) is applied on pre-processed signal to estimate the epochs. The proposed DPI algorithm is validated using six large databases which provide simultaneous EGG recordings. Creaky and singing voice samples are also analyzed. The algorithm has been tested for its robustness in the presence of additive white and babble noise and on simulated telephone quality speech. The performance of the DPI algorithm is found to be comparable or better than five state-of-the-art techniques for the experiments considered.

Data-driven voice source waveform analysis and synthesis

A data-driven approach is introduced for studying, analyzing and processing the voice source signal. Existing approaches parameterize the voice source signal by using models that are motivated, for example, by a physical model or function-fitting. Such parameterization is often difficult to achieve and it produces a poor approximation to a large variety of real voice source waveforms of the human voice. This paper presents a novel data-driven approach to analyze different types of voice source waveforms using principal component analysis and Gaussian mixture modeling. This approach models certain voice source features that many other approaches fail to model. Prototype voice source waveforms are obtained from each mixture component and analyzed with respect to speaker, phone and pitch. An analysis/synthesis scheme was set up to demonstrate the effectiveness of the method. Compression of the proposed voice source by discarding 75% of the features yields a segmental signal-to-reconstruction error ratio of 13 dB and a Bark spectral distortion of 0.14.

Nasal vibration spectra in vowels

An accelerometer microphone attached to the upper part of the nose for measuring pathological nasality has been applied. The advantage of this technique is the easy application, the minor degree of invasiveness, and the minimal speaker distraction. The analysis of the obtained signal has been subject to discussion. Some researchers judged from signal intensity whether speech was produced on a hypernasal level or not. Others calculated the intensity ratio of the accelerometer signal to a voice source signal to eliminate intensity variation. The analysis of the spectral pattern of the accelerometer signal is suggested here. Spectral information in such a signal is a good indicator to reflect the velar port opening, which results in nasal resonance and nasal vibration at the nasal surface. Different vowels induce a different degree of velar opening, which does not necessarily reflect the degree of perceived nasality and measured acoustic resonance. This is due to different nasal and oral impedance of the vowels. Therefore, the spectral patterns of individual vowels in the accelerometer signal are described here. Simultaneously, a fiberendoscopic film has been recorded, showing the velar opening to confirm visually the presence or absence of velopharyngeal coupling.

Acoustic characteristics of voice source for pathological voices: Some case studies.

The aim of this study is to examine the voice source signal of pathological voices. Pathological cases of vocal nodules, polyps, paralysis, malignancy, and laryngitis have been investigated. Some pre‐ and postoperative cases have also been studied. The sustained vowel /a/ of the subject has been recorded. The acoustic signal has been analyzed using an inverse filtering technique to obtain the voice source signal (derivative of glottal air flow). The voice source signal is displayed and marked for detailed examination. Acoustic recording, analysis, and marking is done using a VAGHMI system. For each type of pathology, certain consistent features deviating significantly from the normal cases have been noted in the voice source signal. For example, occurrence of double peak during the open phase, prolongation of the interval near glottal air flow peak, absence of clearly defined negative peak near epoch, etc. Interpretation of the voice source signal in terms of the vocal fold vibratory pattern and pathology has been attempted. Such an interpretation is validated against direct laryngoscopic observation where ever available. This on‐going study suggests the possibility of differential diagnosis of the vocal fold pathology by acoustic analysis of vowel sound.

Glottal inverse filtering by joint estimation of an AR system with a linear input model

Glottal inverse filtering is a technique by which the flow past the time varying glottal constriction is estimated by a filtering operation on the acoustic signal in human speech. The filtering operation removes the effects of the vocal tract resonances to reveal the underlying voice source signal. A linear model of a glottal pulse waveform is described along with a procedure for jointly determining an AR model of the vocal tract response together with the parameters of the glottal pulse model. This technique is applied to inverse filtering in human speech and the results are compared to covariance method LPC analysis.