Speech Segmentation using Regression Fusion of Boundary Predictions

Iosif Mporas , Todor Ganchev , Nikos Fakotakis

In the present work we study the appropriateness of a number of linear and non-linear regression methods employed on the task of combining multiple phonetic boundary predictions. The proposed fusion schemes are independent of the implementation of the individual segmentation engines as well as from their number. In order to illustrate the practical significance of the proposed approach, we employ 112 speech segmentation engines based on hidden Markov models (HMMs). These engines differ in the setup of the HMMs as well as in the speech parameterization techniques they employ. In the experimental evaluation we firstly evaluate the performance of various recent speech features, which have not been tested on the speech segmentation task yet. Secondly, we evaluate the performance of several new fusion schemes for phonetic boundary predictions and finally we contrast them to some recently reported methods. Throughout this comparison, on the established for the phonetic segmentation task TIMIT database, we demonstrate that the support vector regression scheme is capable of achieving more accurate predictions, when compared to other fusion schemes reported so far.

Mots clés: speech segmentation

Iosif Mporas, Todor Ganchev, Nikos Fakotakis. Speech Segmentation using Regression Fusion of Boundary Predictions. Computer Speech and Language, 2009, 24 (2), pp.273. ⟨10.1016/j.csl.2009.04.004⟩. ⟨hal-00593885⟩