Analysis of speaker recognition methodologies and the influence of kinetic changes to automatically detect Parkinson's Disease

Moro-Velázquez, Laureano; García, Jorge Andrés Gómez; Godino-Llorente, Juan Ignacio; Villalba, Jesús; Orozco-Arroyave, Juan Rafael; Dehak, Najim

doi:10.1016/j.asoc.2017.11.001

Cited by 80 publications

(43 citation statements)

References 51 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The obtained baseline results are in the range of those obtained in [10], except for the German database. These differences can be attributable to the use of Albayzin as UBM.…”

Section: Fusion Of Pairs Of Scores From Gmm-ubm With Allophonic DIsupporting

confidence: 57%

“…At this stage, all four parkinsonian speech databases were trained and tested separately following a k-folds crossvalidation scheme, with k=11. These trials were performed using the configurations leading to best results in [10]: Rasta-PLP+∆+∆∆ with number of Rasta-PLP coefficients ranging from 10 to 20 in steps of 2 and 5 coefficients in the FIR filter to calculate derivatives, obtained from 15 ms frames with 50% of overlapping. As the UBM database is sampled at 16kHz, the four databases were downsampled to this frequency.…”

Section: A Baseline: Gmm-ubmmentioning

confidence: 99%

See 1 more Smart Citation

Study of the Automatic Detection of Parkison’s Disease Based on Speaker Recognition Technologies and Allophonic Distillation

Moro-Velázquez

García

Godino-Llorente

et al. 2018

2018 40th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC)

View full text Add to dashboard Cite

The use of new tools to detect Parkinson´s Disease (PD) from speech articulatory movements can have a considerable impact in the diagnosis of patients. In this study, a novel approach involving speaker recognition techniques with allophonic distillation is proposed and tested separately in four parkinsonian speech databases (205 patients and 186 controls in total). The results of applying this new scheme in the databases provides up to 94% of accuracy in the automatic detection of PD and improvements up to 9% respect to baseline techniques. Results not only point towards the importance of the segmentation of the speech for the differentiation of parkinsonian and control speakers but confirm previous findings about the relevance of plosives and fricatives in the detection of parkinsonian dysarthria.

show abstract

“…The obtained baseline results are in the range of those obtained in [10], except for the German database. These differences can be attributable to the use of Albayzin as UBM.…”

Section: Fusion Of Pairs Of Scores From Gmm-ubm With Allophonic DIsupporting

confidence: 57%

Section: A Baseline: Gmm-ubmmentioning

confidence: 99%

Study of the Automatic Detection of Parkison’s Disease Based on Speaker Recognition Technologies and Allophonic Distillation

Moro-Velázquez

García

Godino-Llorente

et al. 2018

2018 40th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC)

View full text Add to dashboard Cite

show abstract

“…Many researches have also been used on speech characteristics for the diagnosis of Parkinson's disease such as PLP, MFCC and Rasta-PLP [10][11][12][13][14]. Savitha S. Upadhyaa et al [10] worked on the detection of Parkinson's disease from the extraction of MFCCs using the multitaper Thomson windowing technique.…”

Section: Introductionmentioning

confidence: 99%

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

Drissi¹,

Soumaya²,

Nsiri³

et al. 2019

ijacsa

View full text Add to dashboard Cite

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 3 rd scale (Daubechie, Symlet, ReverseBior or BiorSpline) combination-MFCC-SVM.

show abstract

“…For recognition process, probabilistic LDA method is used in several kinds of literature. The researchers have focused the speaker verification task with DNN and contribute there result to the classification process [7]. Comparison of standard UBM/i-vector framework with speaker recognition models like universal background model (UBM) and Gaussian mixture model (GMM) gives suitable result for speaker recognition [8].…”

Section: Introductionmentioning

confidence: 99%

Acoustic Feature Extraction and Optimized Neural Network Based Classification for Speaker Recognition

Subhashini¹,

Ram²,

Rao³

2019

IJITEE

View full text Add to dashboard Cite

Identifying the person from his or her voice characteristics is an essential trait for human interaction. Automatic speaker recognition (ASR) systems are developed to find the identity of the speaker in the field of forensics, business interactions and law enforcement. It can be achieved by extracting prosodic, linguistic, and acoustic speech characteristics. Furthermore optimized neural network based approaches are reviewed to classify the extracted features. In this paper, literatures are surveyed on recognition of speaker through the neural network using an optimization algorithm that has developed from the previous years for ASR systems. We deliberate different characteristics of ASR arrangements, containing features, neural network based classification, performance metrics and standard evaluation data sets. ASR system is discussed in two parts. The first part illustrates different feature extraction techniques and the second part involves the classification approaches which identify the speaker. We accomplish this evaluation through a comparative analysis of various recognition of speaker approaches and compare the results of the same

show abstract

Analysis of speaker recognition methodologies and the influence of kinetic changes to automatically detect Parkinson's Disease

Cited by 80 publications

References 51 publications

Study of the Automatic Detection of Parkison’s Disease Based on Speaker Recognition Technologies and Allophonic Distillation

Study of the Automatic Detection of Parkison’s Disease Based on Speaker Recognition Technologies and Allophonic Distillation

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

Acoustic Feature Extraction and Optimized Neural Network Based Classification for Speaker Recognition

Contact Info

Product

Resources

About