pam@uhccux.uhcc.hawaii.edu (.) (01/09/89)
In article <6890> pardo@cs.washington.edu (David Keppel) writes: | >>>Good speech recognition hardware can't be more than 5 or 10 | >>>years away, can it? | >>>-Peter Schachte | diamond@csl.sony.JUNET (Norman Diamond) writes: | >Well, in 1956, in an advertisement on the back cover of Scientific | >American, speech recognition equipment was only 4 years away. | And 1989 - (1956 + 4) = *29* years. Now we're down to only 5 or 10. | Say, that's progress! -- It's here! -- But how come there are no Speech Recognition boards for the Macintosh? There are plenty for IBM PC's. Both Speaker Dependent (you have to teach them your voice first) and Speaker Independent (they'll recognize anybody's voice, but only recognize a limited vocabulary of words) Here are some speech recognition products that are available today for IBM's: VOTAN Voice Card: Using continuous recognition, vocabulary words may be spoken in a natural, conversational flow withourt having to pause between the words. The board recognizezs words from a user's library of word templates. The Votan Voice Card will hold approximately 300 continuous templates and 300 isolated templates Through the use of vocabulary swapping the vocabulary size is constrained only by the computer's memory. The Votan Voice Card has been field tested at 100 decibels with 99% accuracy. The card can operate in any language. [Votan, 4487 Technology Drive, Fremont, CA 94538 (415) 490-7600] Voice Control Systems VCS1000 is a multifunctional speaker-independent voice recognition module for IBM pc's. The VCS 1000 comes with specialized vocabularies such as the "Voice Director" module which allows the recognition of 41 control words such as On, Off, Begin, Stop, Faster, Slower, Left, Right, Up, Down, Forward, Backward, 1,2,3,4,5,6,7,8,9 and zero). This module is useful for factory machine control. Other modules are available and it is also possible to connect the VCS 1000 to the telephone line for speaker independent voice recognition over public switched telephone network. [Voice Control Systems, 14140 Midway Rd Ste 100, Dallas, TX 75244 214-386-5555] Covox, Inc Voice Master PC Digitizer with Voice Recognition is capable of recognizing 64 user-trained words or phrases. User can control the vocabulary syntax and recognition accureacy. Recognition voice templates can be saved to disk. Speech recording and recognition software can be used together for a two way verbal exchange with your computer. [Covox, Inc, 675 Conger St, eugene, Oregon 97402 (503)-342-1271] Speech Systems Incorporated Speech Input Development Syste allows recognition of continuous speech. Not only words, but also phrases and complete sentences. SSI's Speach Input Development System can recognize speech from a variety of speakers. (i.e. it is Speaker Independent) [Speech Systems Inc, 18356 Oxnard St, Tarzana, CA 91356 (818)-881-0885] DRAGON Systems Inc. VOICESCRIBE voice recognition systems allow an active vocabulary of 1000 user difined words and phrases with 99.5% recognition accuracy. It is language independent and functions equally well in noisy environments. It operates on a speaker dependent basis, or with proper training from a cross-section of people, on a speaker independent basis. [Dragon Systems, Inc 90 Bridge St., Newton, MA 02158 (617) 965-5200] Westinghouse Voice Systems Series 1000 Voice Data Entry System offers Speaker Dependent Voice Recognition for data collection. It also offers text to speech voice synthesis. [Westinghouse Voice Systems, East Park 1, Rodi Rd. Pgh, PA 15235 412-825-3500] Astronics ASTROVOIS Voice Operated Information System offers a vocabulary of 1000 speaker dependent words for industrial inspection systems. It is a standard size PC card and comes with 8 diskettes of software. [Astronics, 12501 Prosperity Dr Ste 250 Silver Sprg, MD 20904 (3012)-680-0880] The Voice Connection InfroVoice VI offers Voice Recognition and Speech Synthesis for the IBM PC/XT/AT and PS/2 Model 30. It is a half-size circuit board that offers voice recognition of 500 words with an accurace better than 98%. [Voice Connection, 17835 Skypark Circle, Ste C, Irvine CA 92714 714-261-2366] *** So where are the Voice Recognition systems for the Mac??? *** - Anybody have experience with any of the above mentioned products? Comments? - Anybody know of any other Voice Recognition products for the MAC or IBM? - How about Voice Response (Touchtone) Systems (ala WATSON)? ---end of forwarded message---
pkahn@meridian.ads.com (Phil Kahn) (01/10/89)
In article <2972@uhccux.uhcc.hawaii.edu> pam@uhccux.UUCP (.) writes: > >-- It's here! -- But how come there are no Speech Recognition boards for the >Macintosh? There are plenty for IBM PC's. Both Speaker Dependent (you have to >teach them your voice first) and Speaker Independent (they'll recognize >anybody's voice, but only recognize a limited vocabulary of words) > Check the 7/19/88 MacWEEK (front page). There are two boards from a company called Articulate Systems (415-549-1013). A 1000 word board (VoiceNavigator1000, $1250) and a 200 word board (VoiceNavigator200, $750). Requires training (i.e., speaker dependent). Clever interface to Mac. Is SCSI and ADB based. Two modes for use: 1) program interrupts when keyword detected, 2) have keyword recognition imitate keyboard or mouse action (so it can be front-ended to most applications). I have no experience with the board whatsoever except for literature and conversations with the company. phil...
clive@drutx.ATT.COM (Clive Steward) (01/10/89)
From article <2972@uhccux.uhcc.hawaii.edu>, by pam@uhccux.uhcc.hawaii.edu (.): Gee, that was a really nice list of the PC voice recognition products. The thing is, I think what people are saying is that this kind of product, while useful, is a far distance indeed from machine recognition of even careful casual connected speech. I played with a '1000 word' system like these about 15 years ago, which someone in the Technical Center had bought at Tektronix, for playing with. Cost $10,000 then, was much larger than a PC card. I taught it some Korean, which it (for sound phonemic reasons) phound (it's a fitting brainwave typo, I don't know where these things come from but they're sometimes fun) easier to distinguish than American. I'd just wanted to know if its sound pattern matching algorithms were ethnocentric. What's nice is that the price and size have gone down. And yes, it would surely be fun to have such a card for a Mac bus. Maybe with the new Mac bus? But do people buy Mac II[x]+ for being driven by simple voice commands? Though it could be wonderful for say a person whose body fails them in certain ways. Tell the windows to move around, things to open, talk letters at least into appearing, with all the formatting and see what you'll get advantages of Mac. Maybe we need a Mac bus to PC bus converter just for running such cards. Shouldn't be so expensive, or hard, as long as super performance wasn't goal. Clive Steward
fmg@homxc.ATT.COM (F.GOLDSTEIN) (01/10/89)
In article <6562@zodiac.UUCP>, pkahn@meridian.ads.com (Phil Kahn) writes: > Check the 7/19/88 MacWEEK (front page). There are two boards from a > company called Articulate Systems (415-549-1013). ... > I have no experience with the board whatsoever except for literature > and conversations with the company. I have seen the board in action (Boston MacWorld Expo). I believe that at that time, it was still under development but it was still very slick. Training the board was a snap and the interface was fairly clean. I have no financial affiliation with the company - one of my good friends is an engineer for Articulate Systems. Floyd -- Floyd Goldstein AT&T Bell Laboratories att!homxc!fmg (or) homxc!fmg@att.com arpa!homxc!fmg (or) homxc!fmg@att.arpa
casseres@Apple.COM (David Casseres) (01/11/89)
In article <6562@zodiac.UUCP> pkahn@ads.com (Phil Kahn) writes: >In article <2972@uhccux.uhcc.hawaii.edu> pam@uhccux.UUCP (.) writes: >>... how come there are no Speech Recognition boards for the Macintosh? >Check the 7/19/88 MacWEEK (front page). There are two boards from a >company called Articulate Systems (415-549-1013). A 1000 word board >(VoiceNavigator1000, $1250) and a 200 word board (VoiceNavigator200, $750). Just a point of information: These aren't boards, they're boxes. Being SCSI peripherals, they can be used with any Mac that has a SCSI port, i.e. Mac Plus or later. David Casseres
vincent@ditsyda.oz (David A. Vincent) (01/11/89)
in article <2972@uhccux.uhcc.hawaii.edu>, pam@uhccux.uhcc.hawaii.edu (.) says: > Xref: ditsyda comp.sys.mac:17740 comp.cog-eng:609 sci.lang:22 > > > In article <6890> pardo@cs.washington.edu (David Keppel) writes: > | >>>Good speech recognition hardware can't be more than 5 or 10 **** > | >>>years away, can it? > | >>>-Peter Schachte > [...] > > -- It's here! -- But how come there are no Speech Recognition boards for the No, it is not here. > Macintosh? There are plenty for IBM PC's. Both Speaker Dependent (you have to > teach them your voice first) and Speaker Independent (they'll recognize > anybody's voice, but only recognize a limited vocabulary of words) This kind of stuff has been around for a long time, even for the Apple II. (Ten years?) It does not constitute good speech recognition. Also, I doubt that the so-called 'speaker independent' systems mentioned above will really recognize *anybody's* voice. What about people speaking with strong accents? Or in perfect 'english' but over background noise? > > Here are some speech recognition products that are available today for IBM's: > [...] > > *** So where are the Voice Recognition systems for the Mac??? *** Yes, where? But, by the way, what is voice (as opposed to speech) recognition? (Or is there no difference? In normal discussion, 'voice' is rarely interchangable with 'speech'.) ...D.A.V. David A. Vincent vacation student, CSIRO Division of IT ACSnet: vincent@ditsyda.oz Post: GPO Box 1710-T Phone (w): +61 2 887 9383 Hobart TAS 7001 FidoNet: 3:670/700 Australia ---------------------------------------------------------------- If you don't know how to pronounce it, say it loud! -- Anon. D
srpenndo@uokmax.UUCP (Sean Richard Penndorf) (01/11/89)
In article <2972@uhccux.uhcc.hawaii.edu> pam@uhccux.UUCP (.) writes: => =>In article <6890> pardo@cs.washington.edu (David Keppel) writes: =>| >>>Good speech recognition hardware can't be more than 5 or 10 =>| >>>years away, can it? =>| >>>-Peter Schachte => =>| diamond@csl.sony.JUNET (Norman Diamond) writes: => =>-- It's here! -- But how come there are no Speech Recognition boards for the =>Macintosh? There are plenty for IBM PC's. Both Speaker Dependent (you have to =>teach them your voice first) and Speaker Independent (they'll recognize =>anybody's voice, but only recognize a limited vocabulary of words) Most of the articles I have seen about Mac Speech Recognition are talking about things that are much farther than just Speech Recognition. One company already has the ability to understand speech. However, rather than market what they have so far, they are continuing research. They are attempting to teach their AI software to understand the language of English (later other languages as well) to use it as a dictation word-processor. That is, you only speak to your computer to compose documents or letters. The problems they are having at the moment were the understanding of voice inflections, dialects, and words that sound the same, but are spelled different (hononyms?). Sorry I don't have that article in front of me or I would tell you more. Imagine...Microsoft Word on the Mac where you do not to use a keyboard or a mouse. Now that's Mac user friendly interface! -- Sean 'Longstride' Penndorf !texsun!uokmax!srpenndo . . .----------- GEnie: S.PENNDORF | | `---. srpenndo@uokmax.UUCP `--'LTIMATUM----'OFTWARE
hollombe@ttidca.TTI.COM (The Polymath) (01/12/89)
In article <2972@uhccux.uhcc.hawaii.edu> pam@uhccux.UUCP (.) writes: }Speech Systems Incorporated Speech Input Development Syste allows recognition }of continuous speech. Not only words, but also phrases and complete sentences. }SSI's Speach Input Development System can recognize speech from a variety of }speakers. (i.e. it is Speaker Independent) }[Speech Systems Inc, 18356 Oxnard St, Tarzana, CA 91356 (818)-881-0885] }- Anybody have experience with any of the above mentioned products? Comments? I did some work with the SSI box about a year ago. At that time, what it recognized were "utterances". An utterance could be a word, a phrase, a sentence or almost anything. What it couldn't do was pick individual words out of a multiple word utterance. It could only recognize utterances that were in it's library of utterances. If the sentence "I see a blue box." was in it's library it would recognize it as "Iseeabluebox". If you gave it "I see a red box.", it wouldn't recognize the phrase or, worse, might decide you'd come close enough to "I see a blue box." to recognize it as such. (You could put things in the library like "I see a [blue | red] box.", but that was just a notational convenience). I don't know what developments or improvements SSI has made in the past year, but I assume there've been some, so the above may no longer apply. -- The Polymath (aka: Jerry Hollombe, hollombe@ttidca.tti.com) Illegitimati Nil Citicorp(+)TTI Carborundum 3100 Ocean Park Blvd. (213) 452-9191, x2483 Santa Monica, CA 90405 {csun|philabs|psivax}!ttidca!hollombe
wald-david@CS.YALE.EDU (david wald) (01/13/89)
In article <2209@uokmax.UUCP> srpenndo@uokmax.UUCP (Sean Richard Penndorf) writes: >In article <2972@uhccux.uhcc.hawaii.edu> pam@uhccux.UUCP (.) writes: >>In article <6890> pardo@cs.washington.edu (David Keppel) writes: >>>>>>Good speech recognition hardware can't be more than 5 or 10 >>>>>>years away, can it? >>>>>>-Peter Schachte >>-- It's here! -- But how come there are no Speech Recognition boards for the >>Macintosh? There are plenty for IBM PC's. Both Speaker Dependent (you have to >>teach them your voice first) and Speaker Independent (they'll recognize >>anybody's voice, but only recognize a limited vocabulary of words) > [discussion of up'n'coming Mac speech recognition AI.] > >Imagine...Microsoft Word on the Mac where you do not to use a keyboard or a >mouse. Now that's Mac user friendly interface! Not for me. I've never understood people who can vocally dictate a substantial document. My writing style is just not linear. Part of this is a cut-and-paste style, in which sentences and paragraphs get written first and then put into the proper place, and part is the fact that I'm rarely writing just a single paragraph at a time. I suppose it might work if I had the mouse available to move around, and the speech recognition was only used to insert text, but I can't think of a convenient way to move the insertion point using only speech. Besides, for anything except text insertion, you have to have some way of informing the computer that what you are saying is a direction, rather than more text. What's the speech equivalent of a nonprinting character? (It is *not* an unprintable word.) Now, speech recognition plus eye tracking. Maybe. Now, can we put this discussion in a single newsgroup, please? Followups to comp.cog-eng. -David ============================================================================ David Wald wald-david@yale.UUCP waldave@yalevm.bitnet wald-david@cs.yale.edu "A monk, a clone and a ferengi decide to go bowling together..." ============================================================================
merrill@bucasb (John Merrill) (01/13/89)
In article <1029@ditsyda.oz>, vincent@ditsyda (David A. Vincent) writes: > > >in article <2972@uhccux.uhcc.hawaii.edu>, pam@uhccux.uhcc.hawaii.edu (.) says: >> Xref: ditsyda comp.sys.mac:17740 comp.cog-eng:609 sci.lang:22 >> >> >> In article <6890> pardo@cs.washington.edu (David Keppel) writes: >> | >>>Good speech recognition hardware can't be more than 5 or 10 > **** >> | >>>years away, can it? >> | >>>-Peter Schachte >> -- It's here! > >No, it is not here. No, indeed, it is not here. It is still a long ways away, in fact. Fact: There is *one* (or maybe two) speaker independent, continuous speech recognition system *in existence*. There are no commercial systems extant. The one system, K-F Lee's SPHINX system, runs on "several SUN-4's with floating point coprocessor boards"...and ties them all down. Furthermore, although it is not an isolated word system, it can only handle a finite vocabulary in a *very* limited grammar. I have seen the Lincoln Labs derivative of SPHINX in operation. It's only about 50X real-time, and it isn't bad...if you're running in an absolutely silent room. But it most certainly *isn't* continuous, speaker-independent recognition. (But let me say one thing. SPHINX is a major advance in the design and construction of speech recognizers. No, it ain't perfect...but it's orders of magnitudes better than anything that came before. I was absolutely astounded when it was announced; it's so much better than anything else around. Since I've seen it (or, rather, something very much like it), I'm even more impressed. I just can't convey how much of an advance it was over the older systems.) >Also, I doubt that the so-called 'speaker independent' systems >mentioned above will really recognize *anybody's* voice. What about >people speaking with strong accents? Or in perfect 'english' but over >background noise? I haven't seen any of the new generations of recognizers with accented english, but the one I have seen can deal with a variety of speaking tempi and conditions (yelling, noise-in-ears, deafened, etc.) As I said before, it didn't deal well with noisy environments. On the other hand, there is an accumulating body of evidence that problems with background noise can be ameliorated by the use of non-standard representations of the input stream, some of which appear to be better able to extract signal from background. >> *** So where are the Voice Recognition systems for the Mac??? *** > >Yes, where? But, by the way, what is voice (as opposed to speech) >recognition? (Or is there no difference? In normal discussion, >'voice' is rarely interchangable with 'speech'.) There is a difference. Voice recognition is talker identification (at least, in my jargon). It's much easier than speech recognition. (You can replace speaker dependence with text dependence, and then identify the speaker that spoke your fixed text, as opposed to identifying the text spoken by your fixed speaker.) -- John Merrill | ARPA: merrill@bucasb.bu.edu Center for Adaptive Systems | 111 Cummington Street | Boston, Mass. 02215 | Phone: (617) 353-5765
rcbaab@eutrc3.UUCP (Annard Brouwer) (01/13/89)
Why does everybody want a speech recognition device for their Mac (or other computers)? I've had a test lately with a Sun workstation and an edited text on screen with annotation-markers. You could move the annotation-marker by voice and hear the annotation with a voice command. After a few minutes the system couldn't understand what I said, so I had to repeat my command. This was annoying! All these things could have been done faster with a mouse and a keyboard in my opinion... And do u really want to talk all day *gee* then the world will become more noisy then ever! If you think that I'm wrong, then please let it be, I'm just a simple student who likes mice :) Annard.
pds@quintus.uucp (Peter Schachte) (01/17/89)
In article <434@eutrc3.UUCP> rcbaab@eutrc3.UUCP (Annard Brouwer) writes: >Why does everybody want a speech recognition device for their Mac (or >other computers)? Because I can speak much faster than I can type. For editing, I'd want a combination of mouse and speech recognition (point somewhere and say "insert 'point somewhere and say'," and the phrase appears). Or how about pointing with the mouse and saying "move this...," then moving the mouse and saying "here." One reservation I think people have is about composing "blind." Like composing text by talking into a dictation machine. That requires keeping in mind the structure of the thing being composed (sort of like editing with TECO; anyone remember that?) But it wouldn't be like that. As you speak, your words would appear, just as if you typed them, only much faster. I don't think anyone can predict how it would feel to use such a technology until they've tried it. One thing, though: it'd be pretty hard to enter a program vocally. I guess programming languages would have to get more natural language-like. Maybe COBOL would become the programming language of choice :-). -Peter Schachte pds@quintus.uucp ..!sun!quintus!pds
warner@s3snorkel.ARPA (Ken Warner) (01/18/89)
In article <927@quintus.UUCP> pds@quintus.UUCP (Peter Schachte) writes: >In article <434@eutrc3.UUCP> rcbaab@eutrc3.UUCP (Annard Brouwer) writes: >>Why does everybody want a speech recognition device for their Mac (or >>other computers)? > >Because I can speak much faster than I can type. For a while...maybe. I guess that one might have a different viewpoint if one tried to talk faster than he typed all day, everyday, for a year. Then again, maybe not. I get tired of talking a lot faster than I get tired of typing. > For editing, I'd want >a combination of mouse and speech recognition (point somewhere and say >"insert 'point somewhere and say'," and the phrase appears). Or how >about pointing with the mouse and saying "move this...," then moving >the mouse and saying "here." How would you enter the above paragraph? Ken Warner warner@snorkel.UUCP warner@snorkel.arpa
pds@quintus.uucp (Peter Schachte) (01/19/89)
In article <912@scubed.UUCP> warner@s3snorkel.UUCP (Ken Warner) writes: >>a combination of mouse and speech recognition (point somewhere and say >>"insert 'point somewhere and say'," and the phrase appears). Or how >>about pointing with the mouse and saying "move this...," then moving >>the mouse and saying "here." > >How would you enter the above paragraph? [ using speech and a mouse, and no keyboard ] Maybe I wouldn't have; maybe I would have phrased it differently. But suppose I DID want to enter it just like this. I can think of three ways to do it. The first is the best: a REALLY good AI interface that tries to figure out the punctuation and emphasis you meant. Even if it wasn't perfect, it could probably figure out at least the periods and commas, at least most of the time. If that isn't possible, or if it doesn't work perfectly, I could just say the words, and then go back and put in the punctuation. I could select a quotation and say "quote this." A good voice-operated word processor would know that periods and commas (usually) go inside quotes, but colons (usually) go outside, and other such trivia. It should also know that when you put a quotation inside another quotation, you use single quotes for the inner quotation. So I could have done the nested quotations with two commands and two mouse selections. The parenthesis would be similar. It really wouldn't be so hard. I still believe I could have spoken that paragraph more quickly than I typed it, even if I had to go back and correct punctuation. And I wouldn't have had to worry about spelling. Maybe it wouldn't be as useful as I think. I suspect it would be more useful than you think. But the only way to find out is to TRY such a thing. -Peter Schachte pds@quintus.uucp ..!sun!quintus!pds
rda@epistemi.ed.ac.uk (Robert Dale) (01/19/89)
There's some interesting work that's been done at the University of Dundee by Alan Newell and his colleagues on *simulating* the use of a speech driven word processor (they had someone with a Palantype keyboard -- one of those kbd devices used in court reporting that allows real time transcription -- type what the user said, and then the machine would handle that input). I believe they got quite interesting results -- like it's not at all clear how you could get the interface to such a system to work particularly well. I'm awaiting a paper from Alan on this work -- I'll summarise briefly when it arrives if there's any interest. R -- Robert Dale Phone: +44 31 667 1011 x6470 | University of Edinburgh UUCP: ...!uunet!mcvax!ukc!its63b!epistemi!rda | Centre for Cognitive Science ARPA: rda%epistemi.ed.ac.uk@nss.cs.ucl.ac.uk | 2 Buccleuch Place JANET: rda@uk.ac.ed.epistemi | Edinburgh EH8 9LW Scotland
see1@tank.uchicago.edu (Ellen Keyne Seebacher) (01/20/89)
In article <932@quintus.UUCP> pds@quintus.UUCP (Peter Schachte) writes: > > I could >select a quotation and say "quote this." A good voice-operated word >processor would know that periods and commas (usually) go inside quotes, >but colons (usually) go outside, and other such trivia. It should also >know that when you put a quotation inside another quotation, you use >single quotes for the inner quotation. And would it also know that if the speaker is British (and I presume, but am not sure, Canadian or Australian), the main quotation should use single quotes and the inner quotation double quotes? Or that if the Briton says 'tyre,' she means just that and not "tire"? Would you simply have a Dialect A/Dialect B/Dialect Z switch, or is something more complex needed? -- Ellen Keyne Seebacher University of Chicago Computing Orgzns. see1@tank.uchicago.edu staff.ellen@chip.uchicago.edu
chuck@melmac.harris-atd.com (Chuck Musciano) (01/20/89)
In article <927@quintus.UUCP> pds@quintus.UUCP (Peter Schachte) writes: >In article <434@eutrc3.UUCP> rcbaab@eutrc3.UUCP (Annard Brouwer) writes: >>Why does everybody want a speech recognition device for their Mac (or >>other computers)? > >Because I can speak much faster than I can type. For editing, I'd want >a combination of mouse and speech recognition (point somewhere and say >"insert 'point somewhere and say'," and the phrase appears). Or how >about pointing with the mouse and saying "move this...," then moving >the mouse and saying "here." I wonder about this. Try speaking the preceding paragraph with all the punctuation in place. Those imbedded single and double quotes would be a pain, including the parens and the ellipses. Also, you need some verbal escapes, since the word "insert" is a marker (in the above system) and you need to escape it back to its use as a normal word. The horrors of multi-person offices are too numerous to mention. Not to mention shouting "rm *" into a room crowded with voice recognition terminals. I like quiet when I work, not listening to my office babble to his machine (and vice versa). Chuck Musciano Advanced Technology Department Harris Corporation (407) 727-6131 ARPA: chuck@trantor.harris-atd.com
pds@quintus.uucp (Peter Schachte) (01/21/89)
In article <1487@tank.uchicago.edu> see1@tank.uchicago.edu (Ellen Keyne Seebacher) writes: [ about a speech recognition based word processor ] >And would it also know that if the speaker is British (and I presume, >but am not sure, Canadian or Australian), the main quotation should >use single quotes and the inner quotation double quotes? Sure. And for French, it would use << and >> for quotations. There would have to be different versions for different languages, so why not different versions for different dialects? Or, more likely, you'd plug dialect-specific rules and spellings into a standard "English" word processor. Just like the unix spelling checker that has a "British" mode, and like my Amiga that has different keymaps for different languages/countries. No problem. -Peter Schachte pds@quintus.uucp ..!sun!quintus!pds
lee@uhccux.uhcc.hawaii.edu (Greg Lee) (01/22/89)
From article <1487@tank.uchicago.edu>, by see1@tank.uchicago.edu (Ellen Keyne Seebacher): "... " the Briton says 'tyre,' she means just that and not "tire"? Would you " simply have a Dialect A/Dialect B/Dialect Z switch, or is something more " complex needed? Something more complex, lest she tyre of making corrections. Greg, lee@uhccux.uhcc.hawaii.edu
pds@quintus.uucp (Peter Schachte) (01/25/89)
In article <1444@trantor.harris-atd.com> chuck@trantor.harris-atd.com (Chuck Musciano) writes: >>>Why does everybody want a speech recognition device for their Mac (or >>>other computers)? >>Because I can speak much faster than I can type. >> [ more stuff using lots of single and double quotes and parenthesis ] > I wonder about this. Try speaking the preceding paragraph with all >the punctuation in place. I did. It took about 30 seconds. I didn't pronounce the periods and commas (assuming that a good speech recognition system would infer them from pauses and inflection), but I did pronounce all the other punctuation. I also said "quote" and "end quote" for both kinds of quotes, since a good voice-driven word processor would know the rules about nesting quotations. See my response to the other person who made about the same comment. I just made a comparison. I spoke the last paragraph in about 36 seconds, without rushing. I typed it as fast as I could in 1:44. Try it yourself. Certainly a better typist could type much faster, but 3 times faster? Also note that that paragraph is particularly bad for voice entry. Punctuation slows down voice entry much more than keyboard entry. >Also, you need some verbal >escapes, since the word "insert" is a marker (in the above system) and you >need to escape it back to its use as a normal word. An alternative to this would be having a chord keyboard with just a few keys for handling modes. Holding down one of its keys would mean that spoken words should be interpreted as commands; if the key is up, all words should just be inserted. Another key could be a literal key: when held down, words like "semicolon" would be spelled out, rather than inserting a ';'. Another approach is to plan on the word processor being smart enough to tell from context whether the word makes sense, or must be a command. > The horrors of multi-person offices are too numerous to mention. Not >to mention shouting "rm *" into a room crowded with voice recognition >terminals. The microphones could be insensitive ones (that don't pick up anything more than a foot away, and clipped on your lapel, or one of those boom arrangements that telephone operators use. No problem. >I like quiet when I work, not listening to my office abble to >his machine (and vice versa). Yeah, I agree with you here. I tend to listen to quiet music when I work. It might be distracting if I had to listen to the 5 or 6 people with nearby offices talking to their computers. Maybe it wouldn't work out. But I'd sure like to try it. I'm pretty sure it WOULD be faster. -Peter Schachte pds@quintus.uucp ..!sun!quintus!pds
josh@klaatu.rutgers.edu (J Storrs Hall) (01/26/89)
>> The horrors of multi-person offices are too numerous to mention. Not >>to mention shouting "rm *" into a room crowded with voice recognition >>terminals. > >The microphones could be insensitive ones (that don't pick up anything >more than a foot away, and clipped on your lapel, or one of those boom >arrangements that telephone operators use. No problem. Given that speaker-independent VR is harder than -dependent, I would assume that a one-user voice profile would be common, for reasons both of security and increased accuracy. As a matter of fact, telephone operators work in conditions much more crowded than professional programmers (roughly equivalent to the typical student terminal room). They use voice input devices, i.e. their telephone headsets. Chances are the earphones help in terms of their not distracting one another, but the microphones are unshielded--there are no partitions between stations-- and crosstalk seems to be no problem at all. Modern operator headsets are quite improved over those of even 20 years ago, by the way. They are not particularly onerous to wear for fairly extended periods of time. One could imagine a headset connected to one's workstation that provided one's favorite music as a background to whatever interface was appropriate. --JoSH