Printer Friendly


 LAS VEGAS, Nov. 16 ~PRNewswire~ -- It used to be that talking to your computer meant talking to yourself while your hands and fingers did all the work. Not anymore. IBM (NYSE: IBM) today announced a family of products that allows customers to accomplish real work simply by speaking to their computers.
 Through years of research and a dedication to providing customers with more natural computer interfaces, IBM unveiled the broadest array of speech recognition products available in the industry today. The new products offer a variety of speech recognition capabilities from affordable PC solutions to powerful, leading-edge speech technology, which was developed by IBM's Thomas J. Watson Research Center.
 "Speech is the most natural and efficient means of communications," said James A. Cannavino, IBM vice president and general manager of Personal Systems. "Enabling computers to accept and compute the spoken word can mean tremendous gains in productivity for existing customers and can open up the world of computing to a whole new population of users."
 Speech recognition is the process of accepting human speech into a computer where it is converted to text using sophisticated software algorithms. Text input by voice then is displayed on the monitor, delivered to the application program and stored. Because voice is converted into text, files generated by speech recognition do not take up any more disk space than text input by a keyboard, especially important to PC users with limited hard disk space.
 Our customers will use speech recognition products when ease of use and productivity are important. The workplace segment of the speech recognition market is divided into key areas based on user needs, including general navigation and command, basic business productivity, general dictation and input, advanced function command, data entry and advanced function dictation. IBM meets these needs with a variety of speech recognition solutions that span its entire Personal Systems product line.
 Inexpensive Solutions for the PC User
 The IBM Personal Computer Company brings speech recognition to its customers with two new affordable products that run in a variety of PC- based environments, including DOS, Windows 3.1(See Note B) and OS~2(See Note A) 2.0 (in a DOS window).
 Through a licensing and distribution agreement with Dragon Systems Inc., the PC Company will offer DragonDictate-7K(See Note C) and Dragon Talk-To Plus(See Note C) to its customers.
 Dragon Systems, a leader in speech recognition technology based in Newton, Mass., has worked closely with IBM in Boca Raton, Fla., to develop PC-based speech recognition technology.
 The new products will allow customers to drive applications faster and easier, dictate memos or letters instantaneously without the added step of keyboard data entry and more. Small businesses like insurance agencies that use PC-based forms, pharmacists who type labels or record forms and all non-typist computer users will benefit from using their voice to input data.
 "With speech recognition on the desktop, applications are easier to use, and people are more productive," said Robert J. Corrigan, president of the IBM Personal Computer Company. "Today's announcement brings these benefits to customers at an affordable price."
 DragonDictate-7K, which is designed for general dictation and input applications, allows customers to control many existing applications, such as word processors and spreadsheets, simply by using their voice. Running on any 386SX and above PC equipped with the IBM M-Audio Capture Playback Adapter (M-ACPA), DragonDictate-7K comes ready to use with a microphone and a 100,000-word backup dictionary. DragonDictate-7K runs under DOS, the DOS prompt under Windows 3.1 and OS~2 2.0 (within a DOS window). It automatically adapts to the user's unique voice and allows customers to customize the dictionary with up to 7,000 active words. Available from Dragon Systems or IBM in March 1993, DragonDictate-7K costs $2,295.
 In addition, customers can add simple voice navigation capability to their existing applications with Dragon Talk-To Plus. The program, designed for command and control-type environments, allows customers with a 386SX and above (depending on configuration) to manipulate existing Windows 3.1 applications with simple voice commands, such as "open file," "save file" and "next window." Also available from Dragon Systems and IBM in March 1993, Dragon Talk-To Plus costs $149.
 Highly-Accurate Speech Capability
 IBM today also announced Speech Server Series, a sophisticated, high-performance speech program that uses a RISC System~6000(See Note A) as either a stand-alone workstation or as a server for OS~2 2.0 clients.
 Developed at IBM's Thomas J. Watson Research Center in Yorktown Heights, N.Y., and IBM Vienna Software Development Laboratory, Speech Server Series combines leading-edge software algorithms with high- performance co-processors for sophisticated voice dictation capability. It provides the most accurate speech recognition available. Customers can access 20,000 words in an active vocabulary and can even customize the vocabulary to accept additional, more personalized words. Speech Server accepts dictation at up to 70 words a minute--faster than most computer users can type!
 Speech Server Series was designed to assist organizations producing large quantities of business documentation that encounter delays in transcribing data and for those demanding the highest level of accuracy, such as reporters, healthcare professionals, law enforcement agents, scientists and researchers. It allows customers to create documents in less time by sending commands and text directly to applications by using speech. For example, doctors can read x-rays and record results directly into the computer, or business professionals can dictate memos, letters or even entire reports directly into a word processor without ever touching the keyboard.
 As text is dictated into a headset or lapel microphone, the program is easily trained to accept the speaker's accent. It analyzes the speaker's voice and displays the text in a speech window. The sophisticated statistical language model and high-performance speech co- processors ensure a high degree of accuracy. Words are automatically spelled correctly, and in many instances like-sounding words such as "to," "too" and "two" are successfully distinguished. Before the text is committed to a document, authors can review and edit for accuracy. In addition, the text created can later be formatted, edited and integrated with other applications.
 The IBM Speech Server Series also allows customers to create speech profiles to format data and to interpret commands such as "new paragraph," or "start bold fonts." In addition, it supports multiple languages, including English, French, German, Italian and U.K. English.
 IBM also offers customers its expertise in speech recognition and will assist them in designing new speech enabled applications or integrating speech recognition capability into existing applications.
 IBM Speech Server Series supports up to eight user workstations running either OS~2 or AIX(See Note A) on a local area network (LAN) with a RISC System~6000 server. It also can run on any stand-alone RISC System~6000 workstation equipped with an IBM M-ACPA card. Available in December from IBM and industry remarketers (IR), the AIX Server program costs $6,950, the OS~2 client version costs $695 and the RISC System~6000 speech accelerator card costs $1,900. The AIX stand-alone software costs $2,495 and will be available in May.
 Enable Your Existing Applications with Speech Recognition
 Independent Software Vendors (ISV) will be able in December to develop state-of-the-art speech recognition applications or enhance existing applications quickly and easily with IBM's Continuous Speech Series Developer's Program. IBM's Continuous Speech Series was developed by IBM in conjunction with Carnegie-Mellon University in Pittsburgh, Pa. The developer's program is designed for ISVs who develop OS~2 and RISC System~6000 applications. Through the program, software vendors have access to new speech recognition technology, technical support and sophisticated application program interface (API) to develop applications that accept continuous speech.
 End users will enjoy the natural ease of speaking to their computers and will get faster access to data and applications while keeping their hands free from the keyboard. Continuous speech capability accepts any voice and is an extremely powerful productivity tool for a broad range of applications. For example, a stockbroker could buy shares without ever taking his eyes off the ticker tape or a police officer could check a license plate number without taking his hands off the wheel to write the number down or an air traffic controller could input flight information into the computer without taking his eyes off the radar screen.
 IBM's Continuous Speech Developers Assistance program offers an active vocabulary of 1,000 words from a base vocabulary of 20,000 words. It also offers development tools that can extend the vocabulary to meet the needs of industry-specific applications, speaker-independent operation as well as fast context switching between vocabularies. The developer's tools will be available to software vendors this year for $3,995.
 Note A -- Registered trademark of International Business Machines Corporation
 Note B -- Trademark of Microsoft Corporation
 Note C -- Trademark of Dragon Systems Inc.
 -0- 11~16~92
 ~CONTACT: Tara Sexton of International Business Machines, 702-791-9960 from Nov. 16-20 or 914-642-4662, after Nov. 20~

CO: International Business Machines Corp. ST: New York IN: CPR SU: PDT

WB -- NY045 -- 1226 11~16~92 11:31 EST
COPYRIGHT 1992 PR Newswire Association LLC
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1992 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Publication:PR Newswire
Date:Nov 16, 1992

Related Articles
Is it still IBM versus Macintosh?
Registry Magic Expands Management Team, Appoints Industry Veteran
Torrent Systems Announces Orchestrate Parallel Sorting Component
CallWare, IBM Join Forces to Provide Advanced CTI Solution
CallWare and IBM Successfully Implement Computer Telephony Solutions

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters