Loquendo Launches Expressive TTS

At SYSTEMS 2004, which will take place in Munich, Germany from October 18-22, Loquendo will be unveiling Loquendo TTS Expressive Voices.
 
Developing expressive synthetic speech has been one of Loquendo's objectives for several years; research has covered both language modeling aspects and techniques to make the TTS software manageable, without compromising on the natural qualities achieved through Unit Selection. In the near future, this research effort will make it possible to synthesize text with its preferred style (e.g. emphatic, formal, informal, etc.) and with the emotional tone required (e.g. happy, sad, angry, etc.).
 
Audio 1
Would you like to try it for yourself? Be my guest! Type the text of your choice. Is it difficult? Not at all! It's - how can I say - cool. Or better still: Cool! Good luck. Or should I say: Good Luck!
 
Loquendo's repertoire of "expressive cues" contains conventional figures of speech, like greetings and exclamations ("hello!", "oh no!", 'I'm sorry!"), interjections ("Oh!", "Well!", "Hum"..) and paralinguistic events (e.g. breath, cough, laughter, etc.), which suggest expressive intention (to confirm, doubt, exclaim, thank, etc.).
 
Audio 2
This is an example. Let's assume I am angry. If I were to say: Hey you. Damn it. Hey. This is terrible. Would you believe me? No! But now, I can also say it like this: Hey you! Damn it! Hey! This is terrible! What do you think? It sounds much more human. Good! Speak to you soon!
 
The same elements can appear in several variations in order to obtain the greatest degree of naturalness. The same sentence can be pronounced using different styles and intonations, from neutral to emphatic, from sad to amazed.

Audio 3
Goodness! I sound nearly human. Fantastic. Or would you prefer it if I said: Fantastic! Now, you can choose. Goodbye, or: Goodbye! Excuse me please, or: Excuse me please!

Audio 4
Hello! My name is Stefan. I am Loquendo's new male voice. Now I can sound even more human by using emphatic expressions. For example, I can say: oh all right. But I can also say it like this: Oh all right! Or Oh all right!!
 
The formulas can be typed directly in the text. An interface for prompt editing will suggest the user a repertoire of "expressive cues" available for each voice. The list will be structured according to intuitive linguistic categories, so that appropriate formulas can be found. It will therefore be able to construct messages, dialogues, and narrative that Loquendo's synthetic voices will be able to read.

Audio 5
Stefan:  Hello! Any news?
Katrin:   Gerd M ller will soon be back on the National Team.
Stefan:  Really?
Katrin:   Yes. And so will Beckenbauer.
Stefan:  No way! I can't believe it!
Katrin:   And that's not all! Even Uwe Seeler has been called to play.
Stefan:  Wow! Does that mean we're sure to win this time?
Katrin:   No, of course not!
Stefan:  Damn! ..Hum...Sorry! I meant: what a shame!
Katrin:   But it would be nice though, wouldn't it?
Stefan:  Of course! Have a nice day! Goodbye!