University of Cambridge > Talks.cam > Computer Laboratory Wednesday Seminars > Speech Synthesis at Google

Speech Synthesis at Google

Add to your list(s) Download to your calendar using vCal

If you have a question about this talk, please contact Stephen Clark.

This talk will give an overview of the basics of speech synthesis – from the front-end text-normalisation issues of cleaning up ambiguous text input to the waveform-synthesis backend task of turning a linguistic specification into speech. The two main techniques of unit selection/concatenation and parametric HMM approaches will be covered, with the advantages of each.

The second part of the talk with focus on the applications and rollout of these speech synthesis technologies at Google, dealing with requests from dozens of product teams and balancing requirements and challenges for the future and scaling systems for embedded and server applications and the experience of an acquired startup technology team at Google.

Matt Stuttle is a Research Scientist at Google working in the Speech Synthesis team in London.  Matt was VP Research at Phonetic Arts, a startup focused on speech synthesis technologies for computer games and games development, which was acquired by Google in 2010.  Prior to working at Phonetic Arts, he was at Toshiba’s Cambridge Research Lab doing research and development of speech recognition and synthesis technologies including speech-to-speech translation and in-car and mobile based technologies.  Matt holds a PhD from the Machine Intelligence lab at CUED in speech recognition and was also a Research Associate in statistical spoken dialog systems in the same group.

This talk is part of the Computer Laboratory Wednesday Seminars series.

Tell a friend about this talk:

This talk is included in these lists:

Note that ex-directory lists are not shown.

 

© 2006-2019 Talks.cam, University of Cambridge. Contact Us | Help and Documentation | Privacy and Publicity