Unified communication coffeehouse

A year past and a new UM focus

Christopher Schindler — Tue, 29 May 2007 20:17:00 GMT

A year has passed since my last posting. As you know Microsoft Speech Server was incorporated into Office Communication Server 2007. Since that incorporation our team has evolved into the Unified Communications Application Server team. We are still delivering a great development platform but we've expanded our vision to include unified communications.

I'm limbering up my blogging skills to embrace our new area of focus. I'll share my musings about unified communications, software, and other sundry ideas that take roost in my daily thoughts. Chat with you soon.

ACE conference + singing TTS

Christopher Schindler — Fri, 19 May 2006 00:21:00 GMT

Last week I attended Aculab's ACE conference and participated in a panel discussion. It was a useful trip. I spent time learning about Aculab's VoIP plans and their Prosody products.

Prior to the panel discussion I had a wide-ranging conversation with Judith Markowitz ranging from Speaker Verification(SIV) methods and the VoiceXML forum's SIV planning currently underway. We also discussed the impact of poor application design's impact on speech recognition. And finally ended with classical music and TTS voices singing Frere Jacque; it sounds amazing!

The singing TTS is really fascinating. As some may know part of my distant history was as a professional musician. What is really interesting for me is how prosody would change with singing. The way a musician would shape their mouth to get the right sound would change the phonemes. The phrasing of the music would also change, of couse, the basic prosody. Finally, the lack of breathing is very obvious in the clip I heard. I'm not a linguist or physiologist but I imagine that the consistent air pressure, the volume of the air, the sound volume and the remaining volume of air would really change what we hear. Additionally I hypothesize that a TTS engine would have to make significant changes to handle a realistic synthetic singing voice. A fascinating problem. One other thought: music tuning changes based on the chordal structure of the music -- this I imagine would change the sound of each phoneme as well.

Finally, I participated in a panel discussion, led by Judith Markowitz, with Loquendo, Syntellect, Nuance, and Pronexus. The topic was “Let’s talk about speech” and Judith began with a presentation using the Gartner hype curve and giving her viewpoints on where each of the speech technologies were on that curve. She identified drivers (digitalization of telco networks, customer sat, and growth of the call center), what standards she saw as required (MRCP, SALT and VoiceXML) for platform evaluation and her perspective on Speak Verification and its importance.

Customers had great questions from langauges supported, why MSS chose to support VoiceXML, the value proposition of VoIP, IVR pricing pressures and many others. Thank you to the audience who asked insightful questions.

Cheers!
Chris

Speech-related industry news: searching by voice and voice biometrics

Christopher Schindler — Fri, 14 Apr 2006 00:13:00 GMT

Two interesting articles caught my eye today:

1.) "Google Talks the Talk for Search "
http://www.eweek.com/article2/0,1895,1948889,00.asp

According to the article, a patent has been granted "..to let the human voice command Internet search engines." [eWeek] I have not read the patent but this appears, on the surface, to potentially have a large impact on the speech market depending on how one chooses to define an Internet search engine or Internet search query.

2.) "Say Hello to Voiceprinting"
http://www.techreview.com/BizTech-R&D/wtr_16684,295,p1.html

An article discussing the benefits of voice biometrics.

Cheers!
Chris

Sign up for Microsoft Speech Server 2007 Beta

Christopher Schindler — Thu, 06 Apr 2006 01:00:00 GMT

Hey Everyone,

http://www.microsoft.com/speech/preview/default.mspx

Cheers!

Chris

Announcing Microsoft Speech Server 2007 Beta!

Christopher Schindler — Wed, 05 Apr 2006 19:45:00 GMT

Hi All,

Today we publically announced our plans for Microsoft Speech Server 2007. It's a great feeling to finally take the wraps off the features we've been working on for you. And it will make this blog far more interesting. :)

Check it out: http://www.microsoft.com/presspass/press/2006/apr06/04-05MSS07BetaPR.mspx

Personally, I've been focused on driving the VoiceXML support and supporting VoIP, TDM and TDM/IP hydrids.

I'll be blogging extensively about these areas and I hope we can start a long conversation about the features, the scenarios and your needs.

Cheers!
Chris

New Developer Community

Christopher Schindler — Fri, 17 Mar 2006 02:14:00 GMT

Hi Folks,

Marshall Harrison and Brandon Tyler have put together a very nice developer community site around Speech Server. Check it out!

[EDIT] Thanks to Richard who pointed out that I missed the URL:
http://gotspeech.net/default.aspx

Cheers!
Chris

VoiceXML Port Density?

Christopher Schindler — Wed, 08 Feb 2006 01:54:00 GMT

I heard something surprising the other day. That no VoiceXML platform could support more than 72 ports, w DTMF and some speech recognition, in an all-in-one configuration. Is this true? What's your experience?

Cheers!

Cool contest based on Speech Server

Christopher Schindler — Sat, 04 Feb 2006 02:46:00 GMT

Hi folks,

I just heard about this contest:

www.microsoft.com/speech/evaluation/newsletter/contest.mspx

"Brooktrout is calling all developers to give your application a speech makeover. Create a sample application that targets The Microsoft Speech Application SDK (SASDK) using Visual Studio. The developer with the most creative and interesting application wins a Brooktrout Speech Server Starter Kit and gets bragging rights for creating the best "Code Makeover". Everyone that submits an entry will receive a free long sleeved "Find Your Voice" T-Shirt and all entries will be posted as samples. Your application will help others and spread the word on the power of speech-enabled applications!"

Welcome back, Chris

Christopher Schindler — Sat, 04 Feb 2006 02:42:00 GMT

After emulating a hibernating animal I'm awake and back to blogging. It's been a very rapid and challenging set of months and I can't to share what I've been working on. But that day is not today. :)

Lots of topics have been percolating in my head -- I look forward to sharing them. Talk with you soon.

Be well.

Notify your users...

Christopher Schindler — Fri, 23 Sep 2005 00:32:00 GMT

You've got the coolest speech-enabled app in the market. You believe in your ROI model, you've got business buy-off, and you are laying out your development and deployment plan. As a IT manager/technologist it may be easy to overlook one critical action on your part. Prepare your users. Let them know what changes are coming, when you expect to roll out your new application and what it might be like.

I've heard from several customers that the change from an old IVR system to a speech-enabled IVR can confuse customers. While users like the new interface and find it easier to use dropping the new solution on them, especially if you've only done limited tuning, can result in some interesting user responses. While speech-enabled interfaces are easier to use there is a learning curve for the user. One day, they are listening and hitting touch tones (they know the system). The next day they are expected to speak their response and possibly all the menu architecture/responses have changed. Let them know or risk upsetting your users as they start learning the new system. Your project success, no matter how fantastic your application, hinges on users' success using your application.

Set users' expectations up front. When you release your application they will greet it with open ears and happy hearts. And you’ll get the recognition for a fantastic solution and deployment.

Cheers!
Chris

PS. I'm getting used to the blog format; I'll work on getting more frequent posts out there...

More Katrina Coverage

Christopher Schindler — Thu, 08 Sep 2005 19:53:00 GMT

ITJungle posted this article:

Tech Sector Chips In With Hurricane Relief

Cheers!
Chris

Katrina Support

Christopher Schindler — Thu, 08 Sep 2005 19:31:00 GMT

Every day technologists such as ourselves seek to use technology to improve peoples lives. Over the weekend I was involved in a effort to speech enable an Katrina evacuee application. A number of Microsoft employees, key partners, and service organizations combined to help Katrina survivors and their families and friends. We developed an evacuee tracking application, KatrinaSafe.com, which utilizes a web application and services, .NET client and the Speech Server platform report evacuee status, locate evacuees and notifies people on evacuee status. Just helping reunite one Katrina survivor with one loved one will satisfy our goal.

After assembling a dedicated team of Microsoft technologists and partnering with Intervoice for development and hosting the speech application planning and development started Friday evening. By Tuesday evening the first application was certified for production (this included extensive functional, integration and performance testing) and Intervoice's hosting facility was ready. An amazing effort!

On Monday morning eWeek captured the intensity of the weekend: Microsoft Brings .Net to Katrina Relief Effort

It is incredible to be involved in such a clear application of technology to directly help the people we serve.

Cheers!
Chris

Server or Stand Alone

Christopher Schindler — Wed, 17 Aug 2005 04:32:00 GMT

At SpeechTek a common question was how speech could be embedded on devices. Speech Server is a server based speech recognition, application control, prompting and speech synthesis platform. You can access this powerful platform two ways: 1) telephone only – you call into the Server which creates a voice browser the you interact with and 2) you access the speech application through Internet Explorer on your desktop or your PocketPC/Smartphone. In both cases the speech server does the heavy lifting of speech recognition/speech synthesis/prompt playing and you have to have some connection to the server: either through the telephone only or through Internet access. What this does not provide a stand-alone application on a device that can operate independent of an Internet connection or phone call.

The Speech API or SAPI is the COM-based API method you would use to have stand-alone speech capabilities integrated into your application. The SDK for SAPI is available here: http://www.microsoft.com/speech/download/sdk51. From there you’ll find text to speech engines and speech recognition that you can work with.

Finally, WinFX inlcudes new managed Speech API; using this you can incorporate both speech recognition and speech synthesis. To get this you’ll need to the WinFX SDK for Avalon & Indigo Beta 1 RC 1. Here are some instructions on how to get to this information. http://msdn.microsoft.com/windowsvista/

In the mean time here are some interesting articles to help with SAPI.

“Giving Computers a Voice” -- focuses on using SAPI in C#: http://msdn.microsoft.com/coding4fun/inthebox/TTS-HW/default.aspx
Robert Brown, the Program Manager for the managed Speech API maintains a blog; some interesting reading there: http://blogs.msdn.com/robertbrown/

Cheers!
Chris

VUI Goodbye Anecdote

Christopher Schindler — Tue, 16 Aug 2005 20:16:00 GMT

Hi All,

Following up on our VUI discussion and the "art not yet science" point I made earlier. I was talking with a client who shared this anecdote:

The user had progressed perfectly through the system. The final prompt asked if the user needed anything else to which the user very clearly said "no." The application then launched the final goodbye message saying thank you for using the system, branding message, etc. During this ending message the user must have remembered something and said "Wait, wait, wait...." trying to interrupt the application. However, the application wasn't listening for a barge-in and at the conclusion of the message disconnected the user.

Cheers!
Chris

PS. I've been heads down on a couple of key project for the last two weeks. However, I have been working on a new set of blogs for your consideration. To follow very soon...

Developing without a telephony connection (audience question)

Christopher Schindler — Wed, 03 Aug 2005 18:43:00 GMT

Paul B. kindly asked:

"I'm really interested in speech technology but the biggest hurdle I have seen to actually working with Speech Server is the hardware requirements. I don't have a PBX to hook the server into.

So what I'd love to see is something that could be used to simulate the hardware so that we could then develop apps for fun and profit without having to try to install a PBX system in my house. "

We aim to please. To develop and play with applications you don't need all the telephony hardware. The Microsoft Speech Application SDK (SASDK) has everything you need to develop, debug and simulate calls into your application.

Once you've thoroughly put your application through the wringer perhaps think about taking some next steps...

To begin connecting to the public telephone network you need an interface for Speech Server. It's called a TIM or Telephony Interface Manager. In essence it abstracts the interface to the telephone network and telephony card so you don't have to worry about signalling, call control, etc. Speech Server has certified several TIMs in the marketplace. Then, as Paul wrote, you might need hardware to actually connect to the phone network.

But, there are interesting twists you may consider.

1.) For those of you who happen to have access to a VoIP SIP network you also might consider a SIP TIM; one that connects to your VoIP network or allows you to connect directly with your server using a SIP softphone (software that acts/looks like a phone). We have certified a SIP TIM from Vail Systems and there is another one coming soon from another partner Intervoice.

2.) Consider hosting your application with a 3rd party. A couple of business I'm aware of provide application hosting. They provide the necessary servers and connections to the telephone network including Vail Systems, Intervoice and others.

Cheers!
Chris

This posting is provided "AS IS" with no warranties, and confers no rights.

At SpeechTek

Christopher Schindler — Tue, 02 Aug 2005 16:13:00 GMT

Hi All,

A quick update -- my laptop's hard drive failed on my way here. It should be fixed later today and then I will have several posts ready for your consumption.

Cheers!
Chris

This posting is provided "AS IS" with no warranties, and confers no rights.

SpeechTek

Christopher Schindler — Thu, 21 Jul 2005 19:49:00 GMT

I'll be at SpeechTek in New York from August 1-4. Stop by the Microsoft Speech Server booth and ask for me if you'd like to chat.

I'll be blogging my experiences at SpeechTek so keep watching.

Cheers!
Chris

This posting is provided "AS IS" with no warranties, and confers no rights.

Reference Applications: Need Your Help

Christopher Schindler — Wed, 20 Jul 2005 20:26:00 GMT

One of my tasks is to identify the reference applications that will ship with the next version of Speech Server. Before I prejudice the jury with how we are thinking about reference applications I'd like your feedback. This will be an on-going discussion and I expect to make many posts about this as I hear from you.

Reference applications are sample applications that ship with the product to demonstrate a particular feature set. The code is free for download and review, and we step people through how to create the application in a tutorial format.

Some sample applications that shipped with the v1.x product:

Speech-Enabled Fitch and Mather Stocks Application (FMStocksVoice)
Speech Application SDK -- contained a number of sample applications.

So, to get started, what do you want reference applications to do? How would they best help you? What is the gold standard of reference applications for you? If you were Speech Server, what would be the best reference applications you would develop?

I want to hear from everyone: developers, manager, executives, evaluators, sales, etc. If you don't want to respond publicly on the thread please feel free to contact me directly using the contact link above this post.

Let me say this now: I can't take the covers off our plans for the next release. As we publicly announce details I'll be sure to include them into this conversation and this blog.

Cheers!
Chris

This posting is provided "AS IS" with no warranties, and confers no rights.

VUI Challenge to You

Christopher Schindler — Mon, 18 Jul 2005 22:13:00 GMT

Friday’s Wall Street Journal contained an article by Mr. Wagstaff (available here to wsj.com subscribers) which, I believe, clearly issued a challenge to you. The article discusses the advancements and challenges of voice recognition by focusing on the dictation scenario. Mr. Wagstaff acknowledges that voice recognition is quite good and focuses on the challenge of having voice recognition interact with people. Here is what Mr. Wagstaff had to say:

You see, it turns out that the problem with speech recognition isn't recognizing what you're saying. The problem is interpreting what you want the computer to do beyond that. (Speaking Up for Voice Recognition, Wall Street Journal, July 15, 2005)

Your challenge is summed up in that pair sentences.

Voice User Interface(VUI)design is the art of designing and integrating a speech interface that "anticipate[s] the needs and preferences of the user and conform[s] to the user's mental model of the domain and of spoken language." (Dr. Hura, Heuristics: Lessons in the Art of Automated Conversation) Or, as I like to think about it, it's the “say what?” model. If in a conversation with you about vintage cars I suddenly ask you what time is high tide Seattle you might respond with “say what?” I switched context and you had to quickly interpret what I wanted (you could also say I was a poor conversationalist). It's a challenge of context.

One of the most effective ways we've found of defining context is the core set of scenarios that an application must support. What scenarios are you enabling? What happens if users step outside of those scenarios?

Imagine that you are speech enabling a store locator application. One core scenario you want to enable is: user telephones in, and in three prompt steps is able to find the store closest to them. There could be other scenarios you would like to enable as well: step by step instructions as you travel to the store, a call transfer to an employee of the store for additional details, store item query for a certain item, etc. What are the steps that you take? How would you set the context for the user? What questions would you ask the user that resulted responses that you are prepared to act on? For instance your store application asked: "What time would you like to pick up that item?" Some possible responses: 4 o'clock, in 30 minutes, 1600 hours, this afternoon. What if the response is: "when I'm ready."

My point is that while speech recognition has gotten quite good your challenge is to design a Voice User Interface that enables speech technology to its fullest potential and addresses your key user scenarios.

VUI is a design discipline has entered the mainstream and takes its place as a peer alongside visual design, engineering design, information design and the others. Like the other design disciplines there is technique and art to VUI design. There is an opportunity here: the percentage of people in organizations who are good at VUI and evangelizing speech technologies is very small.

As developers, managers, executives and educators you are tasked with developing your vision for speech enabled applications. Here are some suggested steps:

Identify the right individuals in your organization who will become your leaders in VUI design. Please remember that this expertise spans individual disciplines. For instance, developers might be focused on the tools they needs, the application architectures required to enable the scenarios, error handling, performance, grammars, prompts, etc. Product managers might be focused on the key scenarios, market requirements, business rules, context requirements, etc. What skills do your students, employees and managers need to effectively leverage speech technologies?
Don't bolt-on speech without making VUI design a top priority. For instance, you are speech enabling a preexisting web form -- the appropriate VUI may be different from the form list. For instance, if the form has three text input boxes: size and type. Your speech-enabled application could have users answer in such a way that all three web form boxes could be filled out automatically, such as "I'd like a small pepperoni pizza.” The visual fields could be automatically populated with the values, small and pepperoni. Don't just add-on speech. The chances are very good that you'll miss the right context for your users. Make yours a speech-enabled, not speech-disabled, application by setting the right VUI.
Evaluate your business and service model. How can speech-enabled applications help your organization? Are there obvious opportunities in your organization where speech-enabled applications might save time and money or enhance services?

You are setting the context for your application. What do you want your application to do after the user says hello?

Cheers!
Chris

Additional Resources:

VUI Design
Heuristics: Lessons in the Art of Automated Conversation
Voice User Interface Design - Purpose and Process
Best Practices in Designing Speech User Interfaces
Six Steps for Creating a Speech Recognition Application or Speech-Enabling Your DTMF IVR

Business Value
Speech Server Case Studies

This posting is provided "AS IS" with no warranties, and confers no rights.

Thank you

Christopher Schindler — Sat, 16 Jul 2005 22:40:00 GMT

I heard from a number of you after my welcome message. Thanks very much for welcoming me to the blog-o-sphere. I have enabled comments on my posts which was a popular suggestion. Talk with you soon.

Cheers!
Chris

This posting is provided "AS IS" with no warranties, and confers no rights.

Welcome

Christopher Schindler — Wed, 13 Jul 2005 02:19:00 GMT

Hi All,

A bit about me. I’m a Program Manager in Microsoft Speech Server. I see speech as the next enabler of technology: a means to make interfacing with technology easier and simpler every day. This blog, I hope, will get you thinking about Speech enabling your technology and be a venue for you to teach me what Speech means to you.

Please feel free to contact me and join the blog discussion.

Cheers!
Chris