Slides & video recordings of courses taught at the University of Crete in July 2016.
Part of the Speech Synthesis: Advancements in Modern Speech Synthesis Engines summer school.
Slides
- Text Processing for Speech Synthesis
- Speech Synthesis using Hidden Markov Models
- Evaluating Speech Synthesis
- Hybrid Speech Synthesis
Videos
Tokenisation, normalisation, ...etc
1 hour 32 minutes
Of course, a regression tree is doing the hard work
1 hour 50 minutes
Includes some material on hybrid speech synthesis.
44 minutes (material on hybrid systems starts at 35m30s)