Informatics Forum room 3.33,
10 Crichton Street,
Focus: Speech synthesis
The quality of speech output from HMM synthesis systems is noticibly different from natural speech. By identifying the processes that speech information goes through between natural speech and that output by HMMs, I aim to investigate where these extremely noticable shortcomings in quality observed during HMM speech synthesis are introduced.
- First class BSc (Hons)- Computer science with a year in industry
- Final year project: 'Packet loss compensation in mobile telephony and VoIP'
- University of East Anglia (UEA)