Central Asian Research Index

Al-Farg'oniy avlodlari, 2024, № 4, с. 220-225

AВТОМАТИЧЕСКАЯ СЕГМЕНТАЦИЯ РЕЧИ С ПОМОЩЬЮ МЕТОДОВ КЛАСТЕРИЗАЦИИ

Johongir Urinboev, Mavluda Nugmanova

Аннотация · Abstract

This paper analyzes the effectiveness of modern clustering algorithms for automatic segmentation of speech signals. Segmentation methods based on K-means, fuzzy c-means, and DBSCAN algorithms were applied, and the time boundaries of words were determined using these methods. The obtained results confirm the effectiveness of these algorithms in automatic speech processing. The work also describes an approach to defining precise word boundaries. The results were compared with other methods of speech signal segmentation.

Сведения · Details