Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!wuarchive!psuvax1!hsdndev!bbn.com!nic!news.cs.brandeis.edu!chaos.cs.brandeis.edu!dan From: dan@chaos.cs.brandeis.edu (Dan Schwarz) Newsgroups: comp.sys.mac.hardware Subject: Re: Voice Activated Computer (Articulate Systems) Message-ID: <1991Apr9.033906.15588@news.cs.brandeis.edu> Date: 9 Apr 91 03:39:06 GMT References: <3151@beguine.UUCP> <1991Mar30.132344.41036@eagle.wesleyan.edu> Sender: usenet@news.cs.brandeis.edu Organization: Brandeis University Lines: 93 Hi! I'm a software engineer at Articulate Systems, so I am more than happy to fill you in on this wonderful product :-) Disclaimer: Articulate pays my salary and controls my life. I work on the Voice Navigator development team. So take everything I say with that in mind... In article <1991Mar30.132344.41036@eagle.wesleyan.edu> rcook@eagle.wesleyan.edu writes: >> . . . It's called the Voice Navigator II >> > >You train it to recognize your certain sounds, not necessarily words. In other >words, you can teach it to recognize when you say 'cut' but it won't recognize >when your mother says 'cut' unless she has exactly the same voice as you oryou >teach it to recognize your mother's 'cut'. There didn't seem to be any limit >on how many sounds it could recognize, so you could conceivably have it >recognize the 200 or so words you use most often and then spell out the less >common words. > Better than that - it allows you to create hierarchically structured languages with 200 words on *each level* so with proper structuring, there is no limit to the number of words you can train and have recognized. What do I mean by this? Lets say you have a language with commands that you find in a menu bar: Apple File Edit View Special etc... then on the next level you have commands that you find _in_ the menus: New Open Save Print etc... so this allows you greatly expanded voice recognition and better accuracy. The VN also allows you to have different people train different voice files, so your mom can have her own voice file. This means you can easily set it up to recognize either you or your mom, and switch between the two at will. >Apparently it is connected to the system the same way MacroMaker is. Instead >of being activated by certain keystrokes or menu clicks it is activated by the >sounds you make. It seems to be essentially a front-end for MacroMaker, if you >will. Of course, it doesn't require MacroMaker, but it works the same way, >using sound to generate fake keystrokes or mouse action. In addition, like >MacroMaker, you can have certain sounds associated with only certain programs. > >What would make the product even more attractive would be if it had some sort >of command language, whereby you could dictate actions not expressible as >simple macros. Imagine it being able to interpret this: "go to the finder. >copy the correspondence folder to the internal drive and eject the disk. go to >Word 4. replace attractive with amazing. change font to palatino. change font >to times. save. print on laserwriter. quit. shut down." This might be a little >complicated, and would probably require recognition of text on the screen, but >would be excruciatingly cool. Thanks :-) We can do most of what you ask, right now. The Voice Navigator system comes with a fairly complete set of scripting commands which allow you to control most aspects of the Macintosh interface. You can do things like copy folders, eject disks, start word processors, replace text, change fonts, print, quit, and shut down. A lot of this can be done with the prepackaged languages we provide with each VN, but you'll have to create custom macros on your own (there is a program called Language Maker included which helps you do this.) Future releases of Voice Navigator software will offer much greater functionality. System 7 will provide spiffy new ways of communicating with applications, and we'll be able to exploit that fully. But we can do everything you asked for, right now. > >So, from what I have gathered, the Voice Navigator II is not a SpeakWrite, nor >is it a general speech recognizer like the one Kurzweil makes, but it does >appear to be a solid product. And considering the technology required to >recognize even simple sounds, $600 does not seem unreasonable--it is much more >than a MacRecorder. > >------------------------ >Randall Cook >rcook@eagle.wesleyan.edu >------------------------ ------------------------ You bet. By the way, we do provide all the functionality of products like the MacRecorder, built right in to the Voice Navigator. That is, you can do good quality sound input (audio digitizing up to 22khz) through the Navigator. But The VN is much more than that. And getter better all the time! If you have any input about future products, or how we can improve or present products, I'd love to hear from you. My email address is dan@chaos.cs.brandeis.edu or you can call me at ASI, (617) 876-5236. _Dan Schwarz Software Engineer, Articulate Systems -- | And the men who hold | Dan Schwarz, MB 2926 Brandeis U. | RECYCLE YOUR JUNK| | high places/ Must be | I'NET dan@chaos.cs.brandeis.edu | PEACE IN THE GULF| | the ones to start/ |----------------------------------| tradetapes?mailme| | To mould a new reality/ Closer to the heart ... | FloydRushDead,etc|