The article by Vladimir Fedoseev, Anton Konev, and Alexey Yakimuk from Tomsk State University of Control Systems and Radioelectronics explores the parameters of vowel sounds in Russian and English languages. The study is intended to improve multilingual speech recognition systems that often process signals without prior knowledge of the language. The researchers use a program complex to determine various aspects of the sound, such as the presence of vocalization, frequency and intensity of the main tone, and the ratio of the intensity of harmonics to the intensity of the main tone. The ultimate goal is to develop a model that can recognize and respond to phonetic differences in various languages.
Publication date: 31 Jan 2024
Project Page: Not provided
Paper: https://arxiv.org/pdf/2401.14890
