HMM-based Speech Synthesis of Live Sports Commentaries - Demo

Example 1
Example 2
Example 3
Example 4
Example 5
Example 6
Example 7
Example 8
Base
(manually-checked phonetic transcription)
                               
Loc
(specific information from the local prosody annotation level)
                               
Glob3
(average-voice model adaptation to each speaking style)
                               
Glob3+Loc
(Glob3 + specific information from the local prosody annotation level)
                               

Reference:
B. Picart, S. Brognaux, T. Drugman
"HMM-based Speech Synthesis of Live Sports Commentaries: Integration of a Two-Layer Prosody Annotation"
8th ISCA Speech Synthesis Workshop (SSW8), Barcelona, Spain, 2013

Authors webpage:
Benjamin Picart
Sandrine Brognaux
Thomas Drugman