A non-uniform real-time speech time-scale stretching method

Abstract

An algorithm for non-uniform real-time speech stretching is presented. It provides a combination of typical SOLA algorithm (Synchronous Overlap and Add) with the vowels, consonants and silence detectors. Based on the information about the content and the estimated value of the rate of speech (ROS), the algorithm adapts the scaling factor value. The ability of real-time speech stretching and the resultant quality of voice were analysed. Subjective tests were performed in order to compare the quality of the proposed method with the output of the standard SOLA algorithm. Accuracy of the ROS estimation was assessed to prove its robustness.