An HMM-based system for automatic segmentation and alignment of speech

Abstract

This paper describes an automatic speech alignment system developed for Swedish, which is an extension to earlier work in this area (Sjlander 2001). The system takes as input speech sound files and matching text files containing word-level transcriptions of the speech. A specialised speech recognition system, based on Hidden Markov Models, is applied and a time aligned phone-level transcription is produced