All the Cool Developers use Speech APIs : Speech Recognition

Our Users Are Leading Authorities

Charles Oppermann — Tue, 26 Aug 2008 08:26:53 GMT

Throughout my career at Microsoft, I've eagerly participated in mailing lists, newsgroups, and web forums to engage customers and learn more about their needs and foster direct communication.

One of the better forums for speech recognition is run by Professor Itamar Even-Zohar of Tel Aviv University, where he teaches Culture Research. Itamar has been a long time user of speech recognition and vocal in feedback regarding Windows Speech Recognition. His web site on speech recognition contains useful information on WSR and speech recognition included in Office XP and Office 2003. In particular, his ms-speech forum is invaluable.

Recently when David Pogue of the New York Times wrote about the newest version of NaturallySpeaking, Itamar was quick to write David and set him straight on a few matters, including a plug about Windows Speech Recognition Macros!

David wrote of Itamar, "Clearly, I’ve unearthed the world’s leading authority on speech-recognition foreign-language versions,"

If you read the links I'm providing, you'll see that Professor Even-Zohar is not enamored of all that we do. He's critical of several aspects of WSR and while he "gets it" regarding WSR Macros, he's quick to point out flaws and features.

It's users like this that we need more of; people who are highly experienced and unafraid to share their opinions. The information provided is valuable to me and the rest of the product teams. On the flip side, we have to be careful regarding users expectations. Bending our ear doesn't mean you'll get whatever feature you asked for, and within a particular timeframe.

Oftentimes we'll have more features than time or people available. We have to be very choosy about where to spend our resources. Even things that are a number #1 priority sometimes have to take a backseat to a lesser feature because it was one that we could do in the time or resources available.

Having the feedback from experienced users though help us make the most of the resources we have. We can prioritize better and have confidence that what we're doing will have the greatest impact.

To everyone who writes us at listen, speak and sapitech - we thank you and keep the feedback rolling!

Speech Content in the Windows SDK

Charles Oppermann — Tue, 26 Feb 2008 09:46:00 GMT

I'm happy to announce the availability of the RTM release of the Windows SDK. This release - the first RTM one since Vista - contains the following speech-related items:

Updated: SAPI 5.3 documentation
Updated: System.Speech documentation
Updated: Sample source code

8 C++ projects
3 C# projects
2 sample engines - TTS and SR

New: Grammar Compiler (GC.EXE) tool now part of the tool binaries included in the SDK

The Windows SDK completely replaces the older SAPI 5.1 SDK and supports development on Windows XP, Windows Server 2003, Windows Vista, and Windows Server 2008.

Customers can download this SDK as a DVD image (1,330MB ISO file) from this location:
http://www.microsoft.com/downloads/details.aspx?FamilyId=F26B1AA4-741A-433A-9BE5-FA919850BDBF

Or go through a guided setup process where only the components they need are downloaded. Speech is part of the base install.
http://www.microsoft.com/downloads/details.aspx?FamilyId=E6E1C3DF-A74F-4207-8586-711EBE331CDC

I'm particularly interested in your feedback regarding the Windows SDK as a whole and in particular getting speech information.

InterSpeech 2006

Charles Oppermann — Fri, 22 Sep 2006 08:45:00 GMT

[This posting originally appeared on my personal blog. I'm copying all my speech related blogging to this new MSDN hosted blog. I'll be doing an introduction post soon.]

I'm in Pittsburgh this week, attending the InterSpeech 2006 conference. Actually, I shouldn't say I'm attending it; I'm just staffing the Microsoft booth, giving demonstrations of Windows Speech Recognition. This is an academic conference, mainly for speech scientists and researchers to present their published papers. For example, one of the poster sessions is entitled "A Novel Framework of Text-Independent Speaker Verification Based on Utterance Transform and Iterative Cohort Modeling" which has Microsoft's own Zhengyou Zhang as one of the authors. The poster sessions which remind me of some early science fair projects because it's posted on a wall, with the research data and conclusions neatly shown.

Since Microsoft Research is one of the sponsors, they get a booth in which to demonstrate technology and products. A week ago, the Speech Research Group asked my group, Speech Components, if one of the program managers could come out and give demonstrations. I volunteered. The demos went well, and for the most part were trouble-free. I choose to use the Release Candidate 1 of Windows Vista for the demo machines, because I didn't want to risk problems with an unknown, random build. There was a small issue with the audio gain on the microphone that would set the gain at the maximum after the computer resumed from standby, or the USB headset unplugged and plugged back in. The gain is supposed to be set at 15, so when it went to a 100, recognition accuracy would plummet, but not too badly.

Usually, it was difficult at times to show the correction dialog, used when some phrase was dictated incorrectly. Even when there were hundreds of people milling about the vendor booths, and the ambient noise level very high, the system did very well.

The most often comment was similar to "this is a amazing".