Methods of automatic classification of audio files containing speech in Tatar and in Russian languages

Novel methods are presented for automatic discovery of the speaker's language. The most important feature of the developed algorithms is investigation of local properties of audio file. A list of parameters for des­cription of shape of single wave in audio file is proposed. It is shown, that

Instantaneous Frequency and Detection Dynamics in Speech

Предложен простой метод вычисления мгновенных частот и рассмотрено приложения для задачи разделения источника.

Automatic spoken language identification by digital signal processing methods. Tatar and Russian languages