Developing Acoustic Models for Automatic Speech Recognition of Swedish

This work aims to build robust speaker-independent acoustic models for continuous speech recognition of Swedish using the SpeechDat database.