Current Techniques Speech Technology
Slides (postscript)
Sound samples
- Introductions
-
General Festival introduction (UK English rab)
-
KAL introduction (US English)
-
Spanish example (from 100 years of Solitude)
-
Welsh example
- Text analysis -- examples
-
Boundary prediction: He wanted to go for a drive in.
-
Boundary prediction: He wanted to go for a drive in the country.
-
Homograph disambiguation (pos):
My cat who lives dangerously had nine lives.
-
Homograph disambiguation (CART):
Henry V: Part I Act II Scene XI: Mr X is I believe, V I Lenin,
and not Charles I.
- Mode specific analysis: reading addresses
-
Raw default analysis.
Address mode analysis.
Smith, Bobbie Q, 3337 St Laurence St,
Fort Worth, TX 71611-5484, (817)839-3689
Anderson, W, 445 Sycamore Way NE,
Lincoln, NE 98125-5108, (212)404-9988
- SABLE XML markup language
-
SABLE example (see slide)
- SABLE emphasis examples
-
It will be <emph>rainy</emph> today in Boston.
-
It will be rainy <emph>today</emph> in Boston.
-
It will be rainy today in <emph>Boston</emph>.
- Intonation examples
-
No prosody, fixed F0 and fixed durations.
-
Simple declining F0.
-
"hat" accents on stress syllables.
-
"hat" accents on stress syllables and end tones.
-
Statistically train F0 and durations.
- Waveform generation
-
How not to do word level concatenation.
-
Phoneme concatenation with no modification
-
Standard diphone concatenation.
-
Selection based synthesis (good example)
-
Cluster-based selection synthesis
- Creating diphone databases
-
2 of 1348 diphone prompts
-
Fully automatic diphone based voices
-
After some hand correction
- Cluster synthesis example (copy synthesis)
-
Original
-
Copy synthesized version from phones, durations and F0.
- Telephone dialog examples
- Other synthesis examples
This page is maintained by
Alan W Black awb@cs.cmu.edu
Fluent Speech's demo of
UCSC's Baldi talking head
using
Festival,
with
OGI's voice.