Preface

This website provides a tutorial on how to build acoustic models for automatic speech recognition, forced phonetic alignment, and related applications using the Kaldi Speech Recognition Toolkit.

Acknowledgements

I would like to thank Jack Godfrey, Sanjeev Khudanpur, Paul Smolensky, Yenda Trmal, and Colin Wilson who were integral in creating this tutorial. I am grateful to Jack Godfrey for creating the opportunity for me to learn Kaldi, and to Yenda Trmal and Sanjeev Khudanpur for taking almost an entire day to teach me how to use Kaldi. Yenda Trmal and Paul Smolensky graciously provided comments and revisions on previous drafts of this tutorial. I am also very grateful to Colin Wilson for introducing me to coding and training me as an “apprentice”. All remaining errors are my own.