Talknowledgy Speech technology reference
Chapters 1-7 · speech synthesis, recognition and voice interfaces
Scope
Talknowledgy delivers speech technology products, services, and consulting expertise for multilingual systems, mobile devices, telephone services, voice messaging, and spoken access to dynamic applications.
Consulting and evaluation match customer needs to high-quality speech applications, from multilingual text-to-speech to natural spoken dialogue.
Start here
Read the chapters in order, or jump to the one that answers the question in hand.
- /1/Practice overview4 focus areas
Who the practice is and what fields the work draws on.
- /2/Services9 capabilities
What the practice performs, as service definitions and capability lines.
- /3/Products3 products
Which speech products and voice systems the work is attached to.
- /4/Client engagements16 engagements
Which organisations the consulting work was carried out for, and its scope.
- /5/Publications64 entries
The numbered bibliography, reviews and abstracts, and presentations.
- /6/Appearances7 meetings
Which conferences and meetings the practice appeared at.
- /7/Samplesavailability note
What demonstration material existed and whether it resolves here.
Services reference
| Service | Definition |
|---|---|
| Dialogue and interface design | Spoken and voice user dialogue design for natural information delivery and retrieval. |
| Speech synthesis creation | Engine creation and integration for English varieties, Spanish varieties, French, Italian, German, Swedish, and Japanese. |
| Engine evaluation | Suitability analysis and naturalness testing for concatenative and parametric speech synthesis engines. |
| Custom voices and agents | Customized synthetic voices and speaking animated agents shaped to customer preferences and needs. |
Product specifications
| Product | Specification |
|---|---|
| Apple MacInTalk Pro 2 | High quality synthetic voices released in 1993, including Victoria, Bruce, and Agnes. Victoria's voice was used for Macintosh alert messages. |
| DECtalk software voices | US English male, female, and child software voices from 1995-1997, plus a British English version of DECtalk. |
| Fonix FAAST 4.0 | Fonix female synthetic voice work, including beta availability noted for July 2000. |
Note
The product table carries the release wording published for each product. Fonix FAAST 4.0 is listed at beta availability for July 2000.
Language coverage
| Language | Recorded work |
|---|---|
| English varieties | MacInTalk Pro 2 voices, DECtalk US English male, female and child voices, and a British English version of DECtalk. |
| Spanish varieties | Spanish components for DECtalk text-to-speech, and Latin American Spanish phonetic documentation. |
| German | German components for DECtalk text-to-speech. |
| French | Synthesis engine creation and integration. |
| Italian | Synthesis engine creation and integration. |
| Swedish | Synthesis engine creation and integration. |
| Japanese | Synthesis engine creation and integration. |