Speech segmentation is an essential stage in designing automatic speech recognition systems and one can find several algorithms proposed in the literature. It is a difficult problem, as speech is immensely variable. The aim of the authors' studies was to design an algorithm that could be employed at the stage of automatic speech recognition. This would make it possible to avoid some problems related to speech signal parametrization. Posing the problem in such a way requires the algorithm to be capable of working in real time. The only such algorithm was proposed by Tyagi et al., (2006), and it is a modified version of Brandt's algorithm. The article presents a new algorithm for unsupervised automatic speech signal segmentation. It performs segmentation without access to information about the phonetic content of the utterances, relying exclusively on second-order statistics of a speech signal. The starting point for the proposed method is time-varying Schur coefficients of an innovation adaptive filter. The Schur algorithm is known to be fast, precise, stable and capable of rapidly tracking changes in second order signal statistics. A transfer from one phoneme to another in the speech signal always indicates a change in signal statistics caused by vocal track changes. In order to allow for the properties of human hearing, detection of inter-phoneme boundaries is performed based on statistics defined on the mel spectrum determined from the reflection coefficients. The paper presents the structure of the algorithm, defines its properties, lists parameter values, describes detection efficiency results, and compares them with those for another algorithm. The obtained segmentation results, are satisfactory.

No abstract

No abstract

An adaptive algorithm for vibration signal modeling is proposed in the paper. The aim of the signal processing is to detect the impact signals (shocks) related to damages in rolling element bearings (REB). Damage in the REB may result in cyclic impulsive disturbance in the signal, however they are usually completely masked by the noise. Moreover, impulses may have amplitudes that vary in time due to changes transmission path, load and properties of the noise. Thus, the solution should be an adaptive one. The proposed approach is based on the normalized exact least-square time-variant lattice filter (adaptive Schur filter). It is characterized by an extremely fast start-up performance, an excellent convergence behavior, and a fast parameter tracking capability and make this approach interesting. The method is well-adapted for analysis of the non-stationary time-series, so it seems to be very promising for diagnostics of machines working in time varying load/speed conditions.

No abstract