Archives of Acoustics, 31, 4(S), pp. 205-210, 2006

The application of Kohonen and Multilayer Perceptron Networks in the speech nonfluency analysis

Izabela Szczurowska
Agricultural University of Lublin, Faculty of Agricultural Engineering, Department of Physics, Akademicka 13, 20-950 Lublin
Poland

W. Kuniszyk-Jóźkowiak
Maria Curie-Skłodowska University, Institute of Informatics, Laboratory of Biocybernetics, Pl. Maria Curie-Skłodowska 1, 20-031 Lublin
Poland

E. Smołka
Maria Curie-Skłodowska University, Institute of Informatics, Laboratory of Biocybernetics, Pl. Maria Curie-Skłodowska 1, 20-031 Lublin
Poland

Paper reports the neural network tests on ability of recognition and categorising the nonfluent
and fluent utterance records. 40 of 4-second fragments containing the blockade before
words starting with stop consonants (p, b, t, d, k and g) and including from 1 to 11 stop consonant
repetitions and 40 recordings of the speech of the fluent speakers containing the same
fragments were applied. Two various networks were examined. The first, Self Organizing
Map (Kohonen network), with 21 inputs and 25 neurons in output layer, was used to reduce
the dimension describing the input signals. As a result of the analysis we achieved vectors
consisting of the neurons winning in a particular time point. Those vectors were taken as an
input for the next network that was Multilayer Perceptron. Its various types: with one and two
hidden layers, different kinds and time of learning were examined.
Keywords: neural networks, speech disfluency, Kohonen network, Multilayer Perceptron network, stuttering.
Full Text: PDF


Copyright © Polish Academy of Sciences & Institute of Fundamental Technological Research (IPPT PAN)