Wyniki wyszukiwania - BazTech

1

Deep learning vs feature engineering in the assessment of voice signals for diagnosis in Parkinson’s disease

Majda-Zdancewicz Ewelina, Potulska-Chromik Anna, Jakubowski Jacek, Nojszewska Monika, Kostera-Pruszczyk Anna

Bulletin of the Polish Academy of Sciences. Technical Sciences

|

2021

|

Vol. 69, nr 3

art. no. e137347

EN

Voice acoustic analysis can be a valuable and objective tool supporting the diagnosis of many neurodegenerative diseases, especially in times of distant medical examination during the pandemic. The article compares the application of selected signal processing methods and machine learning algorithms for the taxonomy of acquired speech signals representing the vowel a with prolonged phonation in patients with Parkinson’s disease and healthy subjects. The study was conducted using three different feature engineering techniques for the generation of speech signal features as well as the deep learning approach based on the processing of images involving spectrograms of different time and frequency resolutions. The research utilized real recordings acquired in the Department of Neurology at the Medical University of Warsaw, Poland. The discriminatory ability of feature vectors was evaluated using the SVM technique. The spectrograms were processed by the popular AlexNet convolutional neural network adopted to the binary classification task according to the strategy of transfer learning. The results of numerical experiments have shown different efficiencies of the examined approaches; however, the sensitivity of the best test based on the selected features proposed with respect to biological grounds of voice articulation reached the value of 97% with the specificity no worse than 93%. The results could be further slightly improved thanks to the combination of the selected deep learning and feature engineering algorithms in one stacked ensemble model.

2

Analysis of State-Space Model based Voice Conversion

Sun J., Zhang X.

Przegląd Elektrotechniczny

|

2011

|

R. 87, nr 10

373-376

EN

A new State-Space Model (SSM) based voice conversion method has been proposed recently which outperforms the traditional Gaussian Mixture Model (GMM) method. Although the implementation process of the new method has been elaborated, the theoretical essence of this method has not been analysed clearly. In this paper an exhaustive analysis of the SSM based method is given theoretically and experimentally. Through these analysis, much simpler equivalence form and performance upper bound of the new method are obtained. Finally possible improvements are discussed.

PL

Przedstawiono teoretyczna i eksperymentalną analizę nowego algorytm SSM przetwarzania sygnału mowy.

3

Signal processing of voice in case of patients after stroke

Krzesimowski D., Ciota Z.

Przegląd Elektrotechniczny

|

2010

|

R. 86, nr 11a

129-132

PL

Artykuł dotyczy zmian w parametrach głosu u osób po przebytym udarze mózgu. Przetwarzanie sygnałów może być użyteczne do celów monitorowania postępów hospitalizacji i rehabilitacji. Dokonano nagrań głosu kilkunastu pacjentów, następnie przeprowadzono analizy wszystkich próbek z użyciem różnorodnych algorytmów.

EN

The paper is focused in human voice changes as a stroke result. We suppose that signal processing could be useful to monitor of hospitalisation and rehabilitation progress. The voice of several patients has been recorded; afterwards the analysis of these samples has been performed, by using various algorithms.