Phonetic-acoustic database of highly accented Russian speech
Zalimkhan Nagoev, Irina Aslanbekovna Gurtueva, Kantemir Ch. Bzhikhatlov, M.I. ANCHEKOV · Procedia Computer Science · 2022
In the present work, there is a description for the corpus of highly accented Russian speech. The principles for designing pronunciation dictionary are formulated. And the peculiarities for building on its base sets of allophones, detailed enough for practical use in the field of synthesis of Russian speech and developing speech systems that are resistant to the variety of accents that exist in mountain enclaves. This corpus may be used not only as a method to fix and to monitor peculiarities of the modern Russian. It may also be a useful tool to research problems of perceptive adaptation in oral communication and processes of language convergence and interference of phonetic systems in contact between native speakers of different languages. The proposed approach allows the corpus to be quickly enlarged without significant loss of usability, since the audio data is readily available.