SpeechTEK Europe takes place in only two weeks (26 & 27 May), so if you haven’t already registered, don’t delay any longer .
One part of the show will look at cutting-edge advances in speech technology. Among topics to be discussed are high-definition voice, avatars – the latest in speaking digital assistants, and a new approach to spontaneous speech transcription. One of the session speakers, Ami Moyal, head of the Center for Language Processing at Israel’s Afeka Academic College of Engineering, is particularly excited about the future for spontaneous speech transcription, naming three growth areas.
“Applications include voice mail transcription, in which voice mail messages are transformed into text and sent to subscribers as an SMS, encouraging additional calls and generating more income for the service providers. Inputting SMS messages via voice is another growth area, making SMS accessible to new users and increasing traffic accordingly. And transcribing multimedia audio content for indexing to form the basis of a search application is another important development, improving search engine performance and increasing usage and revenues.”
But he is realistic about the remaining barriers to widespread spontaneous speech adoption. “Spontaneous speech includes a very large vocabulary of words, hesitations and unstructured speech. As such, speech recognition performance is lower than domain-specific applications in which the vocabulary size is lower and language models can be used.”
SpeechTEK Europe Advisory Board member Detlev Artelt, CEO of Aixvox, is enthusiastic about the increased use of speech in real world applications. “As speech goes mainstream we use it in many situations in everyday life,” he says. “A navigation system, for example, understands where we want to go to and can give us instruction by speech. Your telephone understands you, too: just say who you want to call and the machine connects you. It’s so easy because new devices know their owner’s voiceprints and, to ensure data security, they can hide important data from other users.
“Now, the first interactive movie gives the audience its own role by using speech technology to communicate with the protagonist, and help the actor flee the killer,” he continues. “I’m sure we will see many more solutions like this in very near future.”
SpeechTEK Europe takes place at London’s Copthorne Tara Hotel. Registration for both conference and free entry exhibition is at www.speechtek.com/europe, and organizers, Information Today, are offering 40 percent discounts for multiple registrations from a single organization.
