Slashdot Mirror

"I'm sorry. The voice recognition happens inside the device, not on our servers. This wasn't possible 10 years ago, but computers have gotten faster and AI good enough to recognize what you're saying without the assistance of an off-site server. So we never get a copy of what you're saying. After your voice query is recognized, its text version is sent over the Internet if necessary, but we don't keep a record of those either."

Here you go: http://www.nuance.com/dragon/i...

Has been doing everything that Cortana does for decades.

is hte first and most ovbious answer that springs to mind. Not to plug or anything, but Nuance leads the pack with Dragon Naturally Speaking. Version 10 is the best so far (yep, even better than v13), it has a very easy interface and it learns rapidly. http://www.nuance.com/dragon/i...

Gramatically speaking you're right, there is ambiguity. But practically speaking it should be quite clear they're talking about Nuance.

Except that Samsung substantially missed their predicted quarterly earnings for the quarter, whereas Nuance hit the midpoint of their projections.

CereProc did it for Roger Ebert, and other companies such as Nuance and Ivona also offer this service.
I estimate you're about $10k + licencing costs from CBS away from making this a reality.

http://www.cereproc.com/en/services/voicecreation
http://www.nuance.com/for-business/by-solution/custom-voices/index.htm
http://www.ivona.com/en/custom-voice/

There's a gaming speech pack for it that allows you to setup a lot of command types for various games.

http://www.nuance.com/for-individuals/by-product/dragon-for-pc/add-ons/dragon-gaming/index.htm

The cloud offers scalable, redundant and PCI compliant servers

And in order to maintain PCI compliance, you have to restrict access to non-essential personnel, keep your anti-virus updated, encrypt customer data over the network, assign ID's to people with access, yada yada yada. So at least there's that.

Yep, and judging by the buzzword compliant style of verbiage on the website, be prepared to pay out the ass for it. Seems like it would be a bit of a jarring experience. If I'm in my bank app, I would have to switch my brain to the "Nina" context. If I'm in Google Now, I have to switch to that context. I realize this may not be a big deal at all but it is a somewhat disconnected experience. Not only that but app makers can modify Nina in different ways so not only is it different than the standard phone voice input but it's different from itself depending on which app you're in that uses it. Then there's the stigma of talking to your phone anyway since it doesn't even get the "hold it up to your ear" pass that Siri gets. I'll wait and see but I wouldn't expect this to make any huge impact. If anything it will help spur on Apple and Google's efforts.

Dragon is not open source. It is not even multi-platform.

What? Their technology is on multiple platforms and trivially confirmed with google in seconds with queries like: dragon speech mac

WINDOWS: http://www.nuance.com/naturallyspeaking/products/editions/default.asp

MAC: http://www.nuance.com/naturallyspeaking/products/macintosh/for-the-mac.asp

iPhone/iPad: time-limited note recording, but impressive accuracy : http://www.dragonmobileapps.com/

Phone via calling like, as a regular phone: http://jott.com/

Also...

Blackberry: http://appworld.blackberry.com/webstore/content/8108

Dragon is not open source. It is not even multi-platform.

What? Their technology is on multiple platforms and trivially confirmed with google in seconds with queries like: dragon speech mac

WINDOWS: http://www.nuance.com/naturallyspeaking/products/editions/default.asp

MAC: http://www.nuance.com/naturallyspeaking/products/macintosh/for-the-mac.asp

iPhone/iPad: time-limited note recording, but impressive accuracy : http://www.dragonmobileapps.com/

Phone via calling like, as a regular phone: http://jott.com/

Also...

Blackberry: http://appworld.blackberry.com/webstore/content/8108

Dragon is not open source. It is not even multi-platform.

What? Their technology is on multiple platforms and trivially confirmed with google in seconds with queries like: dragon speech mac

WINDOWS: http://www.nuance.com/naturallyspeaking/products/editions/default.asp

MAC: http://www.nuance.com/naturallyspeaking/products/macintosh/for-the-mac.asp

iPhone/iPad: time-limited note recording, but impressive accuracy : http://www.dragonmobileapps.com/

Phone via calling like, as a regular phone: http://jott.com/

Dragon is not open source. It is not even multi-platform.

What? Their technology is on multiple platforms and trivially confirmed with google in seconds with queries like: dragon speech mac

WINDOWS: http://www.nuance.com/naturallyspeaking/products/editions/default.asp

MAC: http://www.nuance.com/naturallyspeaking/products/macintosh/for-the-mac.asp

iPhone/iPad: time-limited note recording, but impressive accuracy : http://www.dragonmobileapps.com/

Phone via calling like, as a regular phone: http://jott.com/

He means Dragon NaturallySpeaking. It is claimed that it is "Up to 99% Accurate". "Up to" means "0% to".

Even if Dragon NaturallySpeaking is 99% accurate, that last 1% is a problem to correct. The software will never make a mistake in spelling. However, it will sometimes substitute similar words that change the meaning of what you intended to say, sometimes in subtle ways.

Dragon NaturallySpeaking has improved a lot since version 7. I don't know whether there were improvements in the recognition engine since version 8.

Sometimes Version 10 Standard is sold at Fry's with rebates that make the total cost $25. However, only the Preferred and more expensive versions allow you to dictate into a handheld recorder for later transcription.

Best algorithm, ever.

How about Dragon Naturally Speaking? It doesn't run on Ubuntu so I haven't tried it, but their video sure is convincing:
http://www.nuance.com/talk/

I will probably start contributing to this FOSS voice-recognition implementation:
http://www.voxforge.org/

You should consider - http://www.nuance.com/pdfconverter/

I am hoping that someone will make a nice personal document management package as free software.

If you use Windows, you can buy this:

http://www.nuance.com/paperport/

The basic features would be:

• Scan in a document (group multiple pages into a single PDF)
• Easily scan a page and insert it into a pre-existing PDF (if you missed a page yesterday, today go back and put it in)
• OCR the documents and provide an index to allow searching
• Provide a really convenient photocopier feature (scan+print)
• Fast and easy. Scan in color, but detect black-and-white and auto-convert to greyscale. Do not pop up any dialogs; when the user clicks on the "Scan!" button, start scanning.
• Also allow dropping in saved HTML pages, OpenOffice.org documents, etc. Manage the user's saved documents, no matter what kind of documents they are.

In a perfect world, the GNOME guys and the KDE guys would both start competing over who can make the slickest product and we all would win.

steveha

Seriously, someone should write something like Nuance Talks for Google Android, especially since the sole means of input is not through a touch screen. Having seen an iPhone, I was struck by two things: 1. It's a huge, ugly-feeling brick of a device. 2. There's only like one button. Either Nuance or Code Factory should port their mobile phone screen reader products to the android platform, or someone should get a group of tallented blind programmers (I don't consider myself tallented) together to write something. Yes, I'm blind, you insensitive clods!

Any Symbian series 60 phone can, in principle, be adapted for blind/low vision users using an application called Talks. The website is http://www.nuance.com/talks/ There is a free trial download you can try.

Get a phone from this list: http://www.nuance.com/talks/phones.asp Then buy this software: http://www.nuance.com/talks/ It's a screen reader for Nokia S60 phones. It is perfect for blind and visually impaired persons. My dad is blind and he's used it for quite a few years now. The supported phones are top of the line. My dad has a Nokia N95, I think, and he really likes that the Talks screen reader is compatible with most of the phone, including email, web browsing, media player, calendar, address book, and the amazing voice commands, which provide a great shortcut for blind users. I highly recommend it. And Nokia phones are the best, IMO.

Get a phone from this list: http://www.nuance.com/talks/phones.asp Then buy this software: http://www.nuance.com/talks/ It's a screen reader for Nokia S60 phones. It is perfect for blind and visually impaired persons. My dad is blind and he's used it for quite a few years now. The supported phones are top of the line. My dad has a Nokia N95, I think, and he really likes that the Talks screen reader is compatible with most of the phone, including email, web browsing, media player, calendar, address book, and the amazing voice commands, which provide a great shortcut for blind users. I highly recommend it. And Nokia phones are the best, IMO.

While I'm all for open source and regarding hardware drivers I wouldn't want it any other way, let's not forget that open source does not have to be pushed around at the application level at the expense of usability. Professional-level applications are critical for the use and expansion of Linux, and proprietary software vendors should be encouraged to develop their software for Linux, not alienated by being badgered to give away their source code. Currently, there is a heated discussion on the Debian list regarding PCB and CAD software availability. One camp (me) is encouraging users to write to software houses and to request that they port their software to Linux, with the other camp rejecting all contact with proprietary software vendors unless it is a demand for the source code. Currently, myself and other engineers cannot use Linux at work because we must run proprietary engineering software, such as Solidworks in my case. For those who want to help, please write to these companies and let them know that we are interested in their software on Linux:

Intuit (Quicken, Quickbooks) http://www.intuit.com/contact/ (requires registration)

Adobe (Photoshop, Flash CS3 Professional, Captivate, Dreamweaver, Studio) http://www.adobe.com/cfusion/mmform/index.cfm?name=wishform

Sony (Vegas Studio) http://www.sonycreativesoftware.com/corporate/contacts.asp

Autodesk (Autocad) http://usa.autodesk.com/adsk/servlet/index?siteID=123112&id=1073074

SolidWorks http://www.solidworks.com/pages/company/SolidWorksOfficeWorldwide.html (requires registration)

Sage (Act!) http://www.act.com/company/contactus/

Nuance (Dragon Naturally Speaking) http://www.nuance.com/help/contact/

hardin-soft (BM-Win Plus (mailing address correction software)) http://www.hardin-soft.com//forms/feedback.html

Daz (Bryce (3D modeling and animation)) http://www.daz3d.com/i.x/support/rnlogin/-/?p_sid=vOwOJN6j&p_accessibility=&p_redirect=&p_lva=&p_sp=&p_li=&p_next_page=std_alp.php (requires registration)

ArenaNet (Guild wars): http://www.arena.net/contact.php

Ironclad Games (Sins of a Solar Empire) http://www.ironcladgames.com/contact.html

Blizzard Entertainment (World of Warcraft) http://us.blizzard.com/support/webform-us.xml?gameId=0

Firzxis (Civilization IV) http://www.firaxis.com/support/

Electronic Arts (lots of games) http://www.info.ea.com/company/company_prlist.php

My personal problem is that I need Solidworks, so for emphasis I'll repeat their address here:
http://www.solidworks.com/pages/company/SolidWorksOfficeWorldwide.html

Please write to these companies and let them know that we need their products on Linux. Copy the list and write to one company a week. Thanks.

It is the problem of verifying a person's identity solely by their voice. It can be used for purposes such as security applications that use a voice print to replace typed passwords and PINs. The voice is then used to authenticate the user.

Yes it matters.

For example, one voice recognition application PowerScribe (from Nuance Dictaphone) works with IExplorer 6.0 only.

Not wit IE 5.5
Not with FireFox
Not with IE 7.0

And it needs Java JRE 1.4.3. Not JRE 1.5, as 1.5 "breaks stuff" (PowerScribe tech support dixit)

This PowerScribe thing uses a whole bunch of wierd ActiveX to launch Java app.

Dictaphone is as a "web application" when it's just a fat client written in Java delivered through IExplroer 6.0+ActiveX from a webserver. IIS webserver.

Ah! you have to wonder what those developers were thinking. ActiveX to launch a Java app? And a Java app limited to an "old" version of JRE?

They were clearly smoking something ilegal.

Peace!

Yes it matters.

For example, one voice recognition application PowerScribe (from Nuance Dictaphone) works with IExplorer 6.0 only.

Not wit IE 5.5
Not with FireFox
Not with IE 7.0

And it needs Java JRE 1.4.3. Not JRE 1.5, as 1.5 "breaks stuff" (PowerScribe tech support dixit)

This PowerScribe thing uses a whole bunch of wierd ActiveX to launch Java app.

Dictaphone is as a "web application" when it's just a fat client written in Java delivered through IExplroer 6.0+ActiveX from a webserver. IIS webserver.

Ah! you have to wonder what those developers were thinking. ActiveX to launch a Java app? And a Java app limited to an "old" version of JRE?

They were clearly smoking something ilegal.

Peace!

but what if you're blind or just have bad eyesight?

Dragon Naturally Speaking! Duh!!!

For work, forget using anything from IBM, forget using the SAP client (SAP GUI) if your work requires it. Forget a *lot* of other essential corporate applications.

Bzzt. Thanks for playing. [PDF] You're right, though. Lots of IBM software is not available for Mac OS X. Like Lotus Notes and ViaVoice.

I'm not going to disagree with you about software availability. However, I think what you'll find is that you may have problems finding a "brand" that you're comfortable with. Everyone's favorite example: AutoCAD. Of course, there are plenty of CAD solutions on the Mac. However, if you must use AutoCAD for some reason, you're out of luck. This is true for games--a particular example of branding. If you're the sort of person who has to play the latest hottest game as soon as it comes out, you're out of luck on the Mac. For example, I got Star Wars: Battlefront for Mac OS X this past Christmas. It shipped in July 2005--about 10 months after the Windows release. So it certainly wasn't the hottest game by the time I got to play it. I still like playing it, though.

Now, how would I convince a consumer to switch to something which is more expensive while lacking both the breadth and width of software and hardware options they currently enjoy? When they ask about upgrades, how would I explain to them that Apple charges rediculous premiums for *any* kind of hardware upgrade? How would I convince my company to switch when just 10% of their essential software even has an equivalent on OSX?

Well, some of your concerns are just plain wrong.

Again, in the consumer realm on the software side, I think you're taking about games. Everything else is covered. In that realm, suggest they buy an Xbox 360 or Wii for the kids to play games with. Mom & Dad will be happier about not having to kick the kids off the computer anyway.

Hardware, I'm a bit lost on. What do you mean "breath of hardware upgrades" for consumers. Are you saying that you can't upgrade memory, hard disks, or video cards on a Mac? Sorry, Macs use industry-standard hard drives. If anything, the problem with Macs are that they tend to be ahead of the curve, so you might have to buy Serial-ATA drives instead of a cheapo ATA/66 hard drive and that will cost you more money. Same with memory--the iMac uses PC2-5300 memory which is more expensive than other types. For example, a 1GB upgrade is about $165. For a Dell XPS200, 1GB of PC2-4200 memory is $120.

From the business software side, I might agree. But, before I do, I'd point out that Macs are cheaper for a business to support than Windows. So, obviously, the more Macs you have, the cheaper it will be. So suppose we ask this question: What job positions require Windows PCs?

Receptionist? That job is mostly about e-mail, etc. All available on Mac. Sales? Again, e-mail, etc. Perhaps some database access for looking up inventory and such, which can be done via the web. Macs have web browsers, so that should be fine. So we could switch Sales. Marketing? Ooh...lots of graphics and such. Plenty of room for Macs in marketing. General Management? Show me a software category that is used by general management that is not available on the Mac.

Engineering? Well, if you're developing Windows software, you need some Windows machines. There are also some great tools for other engineering disciplines which are not available on the Mac. Fair enough. Engineering keeps it Windows machines.

Accounting? There's accounting software for Macs, believe it or not, but I'll let them keep their Windows machines. There are some great accounting solutions for Windows.

You're welcome.

There's voice recognition tools that are designed with that kind of disability in mind. The first I can think of would be Dragon: Naturally Speaking

For those of you in search of voice recognition ware that has already "been developed" you should check out Dragon NaturallySpeaking. I got it for my boss who's paralyzed from the neck down and it works beautifully, making his life easier. Training only took 15 minutes and the accuracy is impressive. It comes with a headset mic but I recommend splurging on the Plantronics CS50-USB wireless headset.

There are lots of OCR programs out there that'll let you convert a PDF/Doc into image files & back again. Probably the only thing you'd lose in the process are hyperlinks embedded in words.

Omnipage is one example.
(I've used their 'lower' end programs PaperPort & PDF Converter)
Acrobat &/or Acrobat Capture can do the same thing.

Other than potential workflow interruptions, I can't see any good reason not to go pdf --> image --> pdf.

Each time speech recognition gets mentioned on /. I see comments about how it doesn't work, or comments on ViaVoice. I am continually amazed at the lack of comments on DNS. DNS8 is a fantastic product with full command and control, auto-punctuation, and almost no training needed. The accuracy is high its hard to demo the slick correction tools; you really need to slur your speach to force it to mis-recognize. Unfortunatly there is no MAC or Lunix support. If you want it, please let the management http://www.nuance.com/naturallyspeaking/support/ know. No work will be done unless there is a buisness case for it.

Also watch for the DNS9 beta starting soon.

Dragon version 8 made major improvements in recognition. The preferred version will read out loud. My wife has neck and shoulder problems, Dragon allows her to use a computer reasonably well. They have ratings for different microphones, I sprung for 5 Dragon usb mic. Doesn't make sense to cheap out after the software's installed. We got an upgrade offer after giving up on Dragon many versions ago. The version 8 release is actually worthwhile.

This sounds ver similar to the built-in voice recognition that macintoshes have had for some time now (not to knock it). Anything you want typed or done can be triggered by a voice command, though it has to be scripted individually.

In terms of Windows, the best that I know of is still Dragon Naturally Speaking, though I strongly recommend pirating it first to decide if it serves your needs. Unfortunately, even with regular training it still gets things wrong with alarming frequency. You have to retouch everything or your mail recipients will think you're an incoherent blogger.

Dragon Naturally Speaking from Nuance is about 75-80% accurate out-of-the-box. It is the other 20-25% that you have to invest the time in to get it to your liking. Even after a few months, you will probably still only reach up to 95% accuracy.
Using it when you have a cold, sore throat or when you have been indulging in your favorite alcoholic beverage can corrupt your voice profile and set you back considerably.

Never let someone else use it under your voice profile.

Will voice rec systems ever be 100% accurate and spearker independant? Maybe, but I don't expect to see it for a long time.

Nuance's (formerly Scansoft) Omnipage

and then a number of smaller players like SimpleOCR

In the open source world, some places to start looking are:

GOCR

and GNU's OCRAD

Both Nuance and Abbyy offer an SDK for OCR integration at a code level which might suit depending on your budget. Certainly the price (probably between $500 and $5000 for a license) represent a good deal if you look at the costs and time it would take to write anything that does serious OCR work yourself.

BTW, if anyone out there knows of any good document layout analysis code available to have a look at, I would be particularly interested. I am looking into document layout analysis for a personal project and although there is a fair bit of academic research available at Citeseer, I actually haven't found much in the way of good sample code that I can use as a starting point for some of my own ideas.

You need a tech update.

Read up on Statistical language models.

aaaand the plug is here: Nuance

ain't quite 100% yet...

And regular speech is?

FYI it's at 97%, give or take a couple. Good enough for TellMe to increase ATT's 800 automation rates from 15% to 70%. Good enough to automate cop cars. It's been a long time coming, and noisy environments are still a challenge, but it's fo real, now.

(Shamelss plug: the really good stuff is running on the engine from Nuance.)

Ooh, a SpeechWorks plug! Okay, let's give equal time to Nuance :)

Nuance Vocalizer Demo

Get an analog SIP gateway, like the one sold by mediatrix.

Then, a VoiceXML Interpreter.

The calls come in the gateway, and get handled by the interpreter, which runs on standard PC hardware. You can configure the interpreter to run different VoiceXML apps based on the caller ID info. You can specify any kind of voicemail app you want in VoiceXML, complete with touch tone and speech recognition.

While you're at it, you can write other vxml apps accessible only to certain people, verified with biometric voiceprint authentication. Here's a scenario: You forgot your housekey. Your electronic garage door opener, however, is hooked up to an X-10 device.

Computer: Hello, would you like to leave a message?
You: This is Joe Shmoe.
Computer: Voiceprint identified. How can I help you?
You: Open the garage.

BTW, the Nuance interpreter comes free with a 2 port license (handles 2 calls simultaneously). Any more than that, and they start charging. The software includes the speech recognition, voiceprint authentication, and voicexml interpreter.

Neat, eh?

(Full disclosure: I have worked with most of these companies).

Telephony-based voice-recognition is going to be the Next Great Thing (tm). The main companies that are involved in this stuff are SpeechWorks, Nuance (both work on the main speech recognition/software stuff), HeyAnita (which works with Sprint), and TellMe.

Disclaimer: I work for Nuance Communications a speech recognition software company

Speech rec works. Talk to Sears, TellMe and Schwab to name a few

This article is not adressing John Doe controlling his PDA with speech commands(although we're getting there). It's talking of Voice browsing the way you browse the web.

We get very high recognition acuracy (90-99% range). Heck, sometimes I listen to .wav where I don't have a clue what the user is saying and the engine is righ on! And with good design you'd be amazed at what voice portals can accomplish.

The trick with voice applications is to design excellent dialogs. Basically, the speech rec will be good if you know what the user is going to say. That means asking the right questions to get the right answers. Then write the grammars that matches. System that are speaker-independant need to have a grammar specified. Basically a list of possible answers the user can give. This network of possibilety can be small or huge (think of diallers with 100k names)

Text-To-Speech also has a bad reputation. We all remember what Dr. SBAITSO used to sound like but beleive me, when the next generation of TTS engine will speak to you you probably won't notice. I heard samples from a few engines and they are unbeleivable.

Take a look at the Voice Web Server. It is available for free (free registration required) on extranet.nuance.com and it works on any Intel PC (yeah, I know...) with a SB16 compatible soundcard

Then come discuss with other users on our discussion forums. I'll be there to answer your questions

Mathieu

Disclaimer: I work for Nuance Communications a speech recognition software company

Speech rec works. Talk to Sears, TellMe and Schwab to name a few

This article is not adressing John Doe controlling his PDA with speech commands(although we're getting there). It's talking of Voice browsing the way you browse the web.

We get very high recognition acuracy (90-99% range). Heck, sometimes I listen to .wav where I don't have a clue what the user is saying and the engine is righ on! And with good design you'd be amazed at what voice portals can accomplish.

The trick with voice applications is to design excellent dialogs. Basically, the speech rec will be good if you know what the user is going to say. That means asking the right questions to get the right answers. Then write the grammars that matches. System that are speaker-independant need to have a grammar specified. Basically a list of possible answers the user can give. This network of possibilety can be small or huge (think of diallers with 100k names)

Text-To-Speech also has a bad reputation. We all remember what Dr. SBAITSO used to sound like but beleive me, when the next generation of TTS engine will speak to you you probably won't notice. I heard samples from a few engines and they are unbeleivable.

Take a look at the Voice Web Server. It is available for free (free registration required) on extranet.nuance.com and it works on any Intel PC (yeah, I know...) with a SB16 compatible soundcard

Then come discuss with other users on our discussion forums. I'll be there to answer your questions

Mathieu

You can look at commercial packages like that offered by VoiceGenie, or Nuance.

Bayonne, suggested earlier, has pretty strict licensing on it for commercial use.

VoiceGenie may be a little young yet with regards to their Linux offering, but it does seem to work ok.

You can check out LinuxTelephony.org, for more ideas.

Caveat with the Dialogic hardware:

• The drivers are closed source and only works on the 2.2.5-14.0 kernel.
• 
  It is dependent on the archaic LiS (Linux Streams) modules.
  Good Luck trying to install security patches or upgrades.
  
  Their hardware pricing is also very strange and counterintuitive.
  Often, More Features != Higher Cost.
  
  There is a new version of the Dialogic drivers coming out, but I've heard they are pretty unstable still, and may not be solid for many more months.