The latest beta releases of Microsoft Speech Server and development platforms add new features in preparation for a full launch in the spring.
Microsoft Speech Server Enters Final Beta
About The Author
Eugenia Loli
Ex-programmer, ex-editor in chief at OSNews.com, now a visual artist/filmmaker.
Follow me on Twitter @EugeniaLoli
I’m speechless. A server to do speech recognition and text-speech. And for some reason it can’t be included in the OS, it needs its own server. I know the mac can speak text, and programs can be used on windows to do speech too. Just there havn’t been any good speech recognition software, so why do we need to write recognition for each piece of software, and not just part of the OS for recognition in all programs. I think someone needs to clearify what MS is doing.
one word: money
with money microsoft can change things, has certain things/aspects in control.
this is the world.
so instead of hitting the keys ctrl+alt+delete you can speak it to the server.. pretty nifty way for rebooting your computer. lol jk. Don’t know, i don’t understand either what is the big deal behind speech recognition software on the server side. i understand an implementation on the client side, but for servers? i don’t see the usefuless of speech, when most servers run headless on my farm.
Speech processing (speech-recognition and text-to-speech) is processor intensive enough to need to be offloaded onto a separate server that is designed to handle these requests in a transactional way in many cases. This would be especially important for high-volume website applications handling lots of traffic.
The principle is similar to how some vendors offer add-in cards to handle (off-load from the server CPU) SSL encryption/decryption processing in a structured way, leaving the server CPU alone to do more important things, like serve web pages.
I guarantee you that you could run Speech Server on the same machine hosting your websites if you really wanted to (for example, MSFT Small Business Server is designed to run Win2003, Exchange, ISA Server, SQL Server and IIS all on the same box).
I guess I could set up a server farm, have it record my voice and translate it into another language for me. Or, I could have it fill out this comment box.
“Microsoft Speech Server supports Speech Application Language Tags (SALT), a proposed speech standard backed by Microsoft.”
I wonder if they are taking XML too far sometimes. I wonder what they are doing that you can’t do with CSS2. Why did they need to invent a markup language for that?
but I am thinking if it could be something like an agent that will talk to you on the phone? Those kind of server?
but I am thinking if it could be something like an agent that will talk to you on the phone? Those kind of server?
Thats what I thought. My assumption was that it allowed one to attach up a regular phone to the network which then interacts with the server and the server handles calls coming from outside the organisation then routes then via the network to the employee.
Now that Gil Bates has cleared up the confusion, it doesn’t sound so interesting after all.
By “regular phone”, I mean a phone designed to work over a lan, not a “normal phone” as in an analogue one used on a normal telephone network.
This is very important, and good news. You were right, ChocalateCheeseCake – it will be used by regular phones, of all types – POTS lines as well as VOIP.
It will allow (with a simple add-on card) for a system to handle voice traffic. This is good for AI geeks like me, who do speech recognition work along with the google development kit. This allows for cool things like the cell-phone call to google – the ultimate expression of a search engine. Imagine being able to call a toll-free number, ask a question and have the choice of hearing the top answers by simple language (i.e. “Give me the first response”). Offloading the recognition to a separate server is good because it allows for clusters to do the recognition, instead of one machine – MS doesn’t currently have a cluster-aware SSI (single system image) aware OS, do they? And no, I’m not being mean – I really don’t know… But yeah – for a guy who hates Windows (it gets in the way for me), I friggin’ love Microsoft’s research initiatives
Peace,
‘rithm
Can’t they just add this to the IIS server? This does seem to be a stupid thing to have one product for. I will bet linux won’t work with it.
I do, and I think a lot of osnews(“Exploring the future of computing”) readers.
I bet that you would too, if the readline was “Open Source Speech Server enters Final Beta”. And no, i dont love ms, and dont hate oss. Quite the opposite.
What’s the point in tying such a feature to IIS, a webserver? Just curious.
Yes, I told you all that Longhorn would have voice recognition, months ago. And did you believe me? Believe me now?
I wonder if they read osnews and /.
This will not be a “free” speech server…*rim shot*