A method for turbulent noise estimation in voiced signals

Mitev, Petar; Hadjitodorov, Stefan

doi:10.1007/bf02344867

Cited by 10 publications

(7 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In these studies, the two-dimensional model is considered as the reference in the predictions comparison. However, since glottal flow is three-dimensional and can develop a turbulent character [17], it is questionable whether twodimensional laminar flow models could provide an appropriate validation for simpler models.…”

Section: Introductionmentioning

confidence: 99%

Experimental validation of quasi-one-dimensional and two-dimensional steady glottal flow models

Cisonni

Hirtum

Luo

et al. 2010

Med Biol Eng Comput

View full text Add to dashboard Cite

Physical modelling of phonation requires a mechanical description of the vocal fold coupled to a description of the flow within the glottis. In this study, an in-vitro set-up, allowing to reproduce flow conditions comparable to those of human glottal flow is used to systematically verify and discuss the relevance of the pressure and flow-rate predictions of several laminar flow models. The obtained results show that all the considered flow models underestimate the measured flow-rates and that flow-rates predicted with the one-dimensional model are most accurate. On the contrary, flow models based on boundary-layer theory and on the two-dimensional numerical resolution of Navier-Stokes equations yield most accurate pressure predictions. The influence of flow separation on the predictions is discussed since these two models can estimate relevant flow separation positions whereas this phenomenon is treated in a simplified ad-hoc way in the one-dimensional flow modelling. Laminar flow models appear to be unsuitable to describe the flow downstream of the glottal constriction. Therefore, the use of flow models taking into account three-dimensional effects as well as turbulence is motivated.

show abstract

Section: Introductionmentioning

confidence: 99%

Experimental validation of quasi-one-dimensional and two-dimensional steady glottal flow models

Cisonni

Hirtum

Luo

et al. 2010

Med Biol Eng Comput

View full text Add to dashboard Cite

show abstract

“…Normalized noise energy (NNE, dB) is a parameter influenced by frequency and amplitude variations and highly sensitive to changes in jitter and shimmer. It was suggested that NNE is extremely helpful in assessing breathy voice [32]. Patients treated with open cordectomy had confirmed incomplete glottic occlusion in more cases than the ones treated with endoscopic laser cordectomy.…”

Section: Discussionmentioning

confidence: 99%

Oncological and Functional Outcome after Surgical Treatment of Early Glottic Carcinoma without Anterior Commissure Involvement

Milovanović

Jotić

Djukić

et al. 2014

BioMed Research International

View full text Add to dashboard Cite

Introduction. Glottic carcinoma can be successfully diagnosed in its early stages and treated with high percentage of success. Organ preservation and optimal functional outcomes could be achieved with wide array of surgical techniques for early glottic cancer, including endoscopic approaches or open laryngeal preserving procedures, making surgery the preferred method of treatment of early glottic carcinoma in the last few years. Material and Methods. Prospective study was done on 59 patients treated for Tis and T1a glottic carcinoma over a one-year time period in a tertiary medical center. Patients were treated with endoscopic laser cordectomy (types II–IV cordectomies according to European Laryngological Society classification of endoscopic cordectomies) and open cordectomy through laryngofissure. Follow-up period was 60 months. Clinical and oncological results were followed postoperatively. Voice quality after the treatment was assessed using multidimensional voice analysis 12 months after the treatment. Results. There were no significant differences between oncological and functional results among two groups of patients, though complications were more frequent in patients treated with open cordectomy. Conclusion. Endoscopic laser surgery should be the first treatment of choice in treatment of early glottic carcinomas, though open approach through laryngofissure should be available for selected cases where anatomical factors present limiting adequate tumor removal.

show abstract

“…After having selected only the speech frames, it is necessary to classify them into turbulent or nonturbulent. To provide this classification we used an algorithm based on a different implementation of the turbulent noise index (TNI) reported in (Mitev, 2000). In the scope of this work, the aim of the TNI parameter was to segment the frames according to the degree of turbulence, not exactly into voiced and unvoiced.…”

Section: Turbulence Segmentationmentioning

confidence: 99%

“…Among the acoustic measures, the most widely used are jitter (changes in pitch period with time) and shimmer (changes in amplitude with time) (Deliyski, 1993), harmonic-to-noise ratio (HNR) (Krom, 1993), cepstral peak prominence (CPP) (Hillenbrand, 1996), glottal to noise excitation ratio (GNE) (Michaelis, 1997), normalized noise energy (NNE) (Kasuya, 1986), soft phonation index (SPI) (Deliyski, 1993), and voice turbulence index (VTI) (Mitev, 2000). Over the years, several systems combining various acoustic measures and different classifiers have been designed.…”

Section: Introductionmentioning

confidence: 99%

“…In an attempt to fill this gap, the aim of the present study is to provide a new point of view concerning the discrimination of acoustic signals using voiced and unvoiced parts. To obtain these unvoiced regions a segmentation algorithm based on an acoustic measure called turbulent noise index (TNI) (Mitev, 2000) is proposed. The classifier is a multilayer perceptron neural network, which uses temporal delays in its inputs.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Pathological Voice Detection Using Turbulent Speech Segments

Perdigão

Neves²,

Sá

2012

Proceedings of the International Conference on Bio-Inspired Systems and Signal Processing

View full text Add to dashboard Cite

Identification of voice pathologies using only the voice signal has a great advantage over the conventional methods, such as laryngoscopy, since they enable a non-invasive diagnosis. The first studies in this area were based on the analysis of sustained vowel sounds. More recently, there are studies that extend the analysis to continuous speech, achieving similar or better results. All these studies use of a pitch detector algorithm to select only the voiced parts of the acoustic signal. However, the existence of a pathology affecting the speaker's vocal folds produces a more irregular vibration pattern and, consequently, a degradation of the voice quality with less voiced segments. Thus, by selecting only clear voiced segments for the classifier, useful pathological information may be disregarded. In this study we propose a new approach that enables the classification of voice pathology by also analyzing the unvoiced information of continuous speech. The signal frames are divided in turbulent/non-turbulent, instead of voice/non-voiced. The results show that useful information is indeed present in turbulent or near unvoiced segments. A comparison with systems that use the entire signal or only the non-turbulent frames shows that the unvoiced or highly turbulent speech segments contain useful pathological information.

show abstract

A method for turbulent noise estimation in voiced signals

Cited by 10 publications

References 15 publications

Experimental validation of quasi-one-dimensional and two-dimensional steady glottal flow models

Experimental validation of quasi-one-dimensional and two-dimensional steady glottal flow models

Oncological and Functional Outcome after Surgical Treatment of Early Glottic Carcinoma without Anterior Commissure Involvement

Pathological Voice Detection Using Turbulent Speech Segments

Contact Info

Product

Resources

About