At Microsoft, we're taking Speech to the mainstream and you can see that with the investments we have been making on this technology in the more recent times since the Speech Application SDK API 1.1 and the GRXML specification tools for the Microsoft-Visual Studio 2003. And, all the most recently the amount of work that is going into Windows Vista. Speech is taking center stage along side many other platforms at Microsoft. The Speech Recognizer (SR) and the Speech Synthesizer engines in Windows Vista are state of the art with many pieces in those two stacks are reengineered to bring the most natural interaction experience out there thus far. The Microsoft Speech Recognition engine thus far prior to Vista was mostly just for the English language. However, with Vista, we are planning on making available at least 8 different languages enabled SR - such as French, German, Chinese, etc right out of the box. And, also working closely with Global ISVs to extending that count to far more. As you can already tell Speech as a platform has never been given this much attention at Microsoft.
So, what is so special about Speech in Vista. For starters and probably my most favorites is the "Speech Enabled Shell". Users are now able to fully interact with their Vista PCs without having to use Human Interface Devices - Mouse or Keyboard. This is awesome step forward especially to our accessibility userbase. If you're or have been a speech developer, you could tell there are really two modes of speech recognition. One, preset language grammar commands and Two, dictation mode. Historically, due to not so matured technological advancements both in software and hardware, the dictation mode has been coined the culprit of putting down speech as a technology and as a means to interact with your PC. However, with the new developments in Speech for Vista, these beariers have been lifted and a sense of true dictation can be experienced by our users. Dictation as one of the two modes is "baked" into the Vista Operating System. What this means for developers is, without having to write a single line of code, speech is automatically enabled into their application dialog boxes. Suppose you had a ::textbox, your users are now able to speak into it with zero-application-developer-coding effort. Your application when run on Vista, is automatically "speech enabled" for dictation mode. How cool is that!
Another break through news for speech application developers like me is the availability of Speech APIs for .Net 2.0. Microsoft is exposing the Speech APIs for .Net 2.0 via the new System.Speech namespace. The System.Speech namespace is going to delivered as one of the components in WinFX::WPF. WinFX is the Microsoft new programming model for .Net that extends the existing featureset of .Net to areas such as Communication (via Windows Communication Foudation - WCF), Presentation (Windows Presentation Foundation - WPF), and Application Workflow (via Windows Workflow Foundation - WinWF). A much debate was done internally at Microsoft as to where to fit Speech into :). It became clear eventually that Speech is a form of presentation and interaction with the user and must therefore be exposed via WPF.
Speech is going to be a huge wave in application development - primarily to bringing higher levels of productivity to your users so they can much intimately interact with their Vista PCs. Speech plays a vital role in "Small Surface" PCs - such as embedded devices running Embedded XP. It will play a huge role in Automobile industry where the Windows Operating System will play an enormous role to improve the productivity of the driver and passengers - both from entertainment and work-productivity standpoint. One of my research projects "iCar-iHome" has been a prime platform for me to enable speech into. Although, the version 1 of iCar is not using Speech namespace, but instead, SAPI 5.0, I am working on it to moving it to System.Speech and leveraging all the benefits it can get from that namespace.
I hope you give the System.Speech a wind and testdrive the platform. If you're interested in getting the latest publicly (at the time of writing this, Feb 06 CTP release bits) available bits, you need to download the WinFX SDK that includes the WPF::Speech SDK and the Runtime that is also needed (if you're only going to be running Speech enabled apps) runtime and sdk in it.
Note:WinFX is scheduled for release along side Vista release (April 2006). However, WinFX is NOT in complete dependency with Vista. Some pieces of WinFX that "light-up" your applications are Vista dependant.
caioNagi