Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Microcap & Penny Stocks : E-COMMERCE by JVWEB, INC. (OTCBB:JVWB) -- Ignore unavailable to you. Want to Upgrade?

To: GC who wrote (429)	3/11/1999 3:29:00 PM
From: GC	Respond to of 767

Saturday, October 17, 1998 The message is: Choose the right words More FP Technology stories By GEOF WHEELWRIGHT For The Financial Post When most people think about speech recognition, recognition of continuous dictated speech comes to mind. This is the most complex and demanding of all speech-related computer technologies, and not even the application with the widest potential benefit. Speech-related computer technologies divide up into three kinds: text-to-speech, command-recognition and dictation-style speech-recognition. A look at each of these in detail reveals the full range of uses of this technology. Text-to-speech technology was pioneered as a way of helping visually disabled people hear a spoken version of computer-generated text. Now, however, it is proving to be invaluable when people want to access computer-generated text, but are unable to read it on a screen at that moment -- for example, someone driving a car who wants to be able to check e-mail. Microsoft and Clarion have developed a system based on Microsoft's Windows CE computer operating system that enables e-mail to arrive via a wireless data connection and be read out using text-to-speech technology. Command-recognition is really a subset of speech-recognition technology, but is far more reliable, requires much less computer-processing power and is much faster than dictation-style speech-recognition. This is because command-recognition enables the system to work with a much smaller set of anticipated responses than a dictation system. The most basic form of command-recognition is used in many automatic-call-response (ACR) and voice-mail systems. These rely on anticipating only the tones that can be generated from a touch-tone telephone system, with preprogrammed responses for each unique set of tones. Voice-based command-recognition provides for only a handful of appropriate voice responses and reacts differently to each. By limiting the number of responses, the command-recognition system can allow for a much wider variation in the tone, accent, pitch and intonation of the user. Voice-command systems require little or no training of the system. This contrasts sharply with dictation systems that often require you to spend an hour or more speaking into the computer, so that it becomes trained to the sound of your voice, and, as a result, can recognize words and phrases more accurately. On a computer, command recognition can be used to start applications quickly, open new on-screen windows, even execute "macros" or mini programs that perform a set of predetermined actions on the user's behalf. When used with a Web browser, common commands such as "up," "down," "back" and "forward" can be recognized quickly and executed. Dictation-style speech-recognition. Continuous speech-recognition has long been the holy grail of speech-recognition researchers, yet it faces a number of inherent technological and sociological barriers. The technological barriers include sound dictionaries large enough to encompass the range of words typically used by most people during dictation. The system must also have enough storage and processing power that it can store the voice patterns of users and match them against appropriate words and phrases. It must be able to handle tough linguistic challenges, such as handling homonyms. In a sentence, such as "Mr. Wright makes the right choice and writes it down," the system would have to distinguish between three different uses of the same sound pattern and select the correct word based entirely on context. This is not a trivial task. Plus, the sociological issues are great. In many offices, users would find themselves feeling uncomfortable talking directly to their computers as they write reports or memos. This potentially robs them of the privacy afforded by a screen turned from prying eyes. Lastly, the notion that computers can talk suggests a disquieting autonomy for these ubiquitous and increasingly powerful devices. Who can forget the evil machinations of the HAL9000 computer in 2001: A Space Odyssey? So far, computers can only think to themselves, "Sorry, Dave. I can't do that." CANOE home \| We welcome your feedback. Copyright © 1999, Canoe Limited Partnership. All rights reserved. Please click here for full copyright terms and restrictions.

To: GC who wrote (429)	3/11/1999 3:32:00 PM
From: GC	Respond to of 767

yes another article Saturday, October 17, 1998 The message is: Choose the right words More FP Technology stories By GEOF WHEELWRIGHT For The Financial Post When most people think about speech recognition, recognition of continuous dictated speech comes to mind. This is the most complex and demanding of all speech-related computer technologies, and not even the application with the widest potential benefit. Speech-related computer technologies divide up into three kinds: text-to-speech, command-recognition and dictation-style speech-recognition. A look at each of these in detail reveals the full range of uses of this technology. Text-to-speech technology was pioneered as a way of helping visually disabled people hear a spoken version of computer-generated text. Now, however, it is proving to be invaluable when people want to access computer-generated text, but are unable to read it on a screen at that moment -- for example, someone driving a car who wants to be able to check e-mail. Microsoft and Clarion have developed a system based on Microsoft's Windows CE computer operating system that enables e-mail to arrive via a wireless data connection and be read out using text-to-speech technology. Command-recognition is really a subset of speech-recognition technology, but is far more reliable, requires much less computer-processing power and is much faster than dictation-style speech-recognition. This is because command-recognition enables the system to work with a much smaller set of anticipated responses than a dictation system. The most basic form of command-recognition is used in many automatic-call-response (ACR) and voice-mail systems. These rely on anticipating only the tones that can be generated from a touch-tone telephone system, with preprogrammed responses for each unique set of tones. Voice-based command-recognition provides for only a handful of appropriate voice responses and reacts differently to each. By limiting the number of responses, the command-recognition system can allow for a much wider variation in the tone, accent, pitch and intonation of the user. Voice-command systems require little or no training of the system. This contrasts sharply with dictation systems that often require you to spend an hour or more speaking into the computer, so that it becomes trained to the sound of your voice, and, as a result, can recognize words and phrases more accurately. On a computer, command recognition can be used to start applications quickly, open new on-screen windows, even execute "macros" or mini programs that perform a set of predetermined actions on the user's behalf. When used with a Web browser, common commands such as "up," "down," "back" and "forward" can be recognized quickly and executed. Dictation-style speech-recognition. Continuous speech-recognition has long been the holy grail of speech-recognition researchers, yet it faces a number of inherent technological and sociological barriers. The technological barriers include sound dictionaries large enough to encompass the range of words typically used by most people during dictation. The system must also have enough storage and processing power that it can store the voice patterns of users and match them against appropriate words and phrases. It must be able to handle tough linguistic challenges, such as handling homonyms. In a sentence, such as "Mr. Wright makes the right choice and writes it down," the system would have to distinguish between three different uses of the same sound pattern and select the correct word based entirely on context. This is not a trivial task. Plus, the sociological issues are great. In many offices, users would find themselves feeling uncomfortable talking directly to their computers as they write reports or memos. This potentially robs them of the privacy afforded by a screen turned from prying eyes. Lastly, the notion that computers can talk suggests a disquieting autonomy for these ubiquitous and increasingly powerful devices. Who can forget the evil machinations of the HAL9000 computer in 2001: A Space Odyssey? So far, computers can only think to themselves, "Sorry, Dave. I can't do that." CANOE home \| We welcome your feedback. Copyright © 1999, Canoe Limited Partnership. All rights reserved. Please click here for full copyright terms and restrictions.