2002 Conference Proceedings

Go to previous article 
Go to next article 
Return to 2002 Table of Contents


TALKING TO YOUR MAC SPEECH RECOGNITION OPTIONS FOR THE MAC OS

Mike Peplow
Adaptive Technology Specialist
Peplowm@missouri.edu 
Ph. 573 882-5657

Adaptive Computing Technology Center
S5E Memorial Union
University of Missouri-Columbia
Columbia, MO 65211

Summary

This presentation will explore the different speech recognition software available for the Macintosh OS 9. x . Focusing on learning curve, usability, cost vs. abilities.

It is truly a shame given the fact that Apple was the first graphical user interface to incorporate speech as part of the operating system. Now, to have everyone that wants to listen to and or speak to their computer running for a Windows-based machine. Why did Apple start the ball rolling only to give up after everyone was hooked?

For too many years Apple tried to work autonomously from the rest of the software developers who were out there writing applications. Apple did this by keeping the Macintosh operating system closed from most of the small software developers, consequently closing themselves off from some of the new wave of products that would be developed. The Macintosh a environment has also been notoriously more difficult to write applications for been the Windows operating system. These two factors have weighed heavily on many software developers only creating their applications for Windows environment. However, there have been daring souls who would strive against convention and still develop applications for the Macintosh operating system. IBM has also taken the opportunity to fill obviously void with probably more of an eye on the profit margin than on the altruistic value it would have.

Individuals who need or preferred to use a Macintosh should no longer feel relegated to use just the keyboard and mouse. Hopefully with more options becoming available every year, the ability to work more productively in the Macintosh environment, particularly for those of us who have mobility impairments will only get better.

Some of the speech recognition products that are available now are IBM's ViaVoice, Mac Speech also has iListen and some of the more advanced options would be running Virtual PC with Dragon NaturallySpeaking. These along with some smaller light weight shareware applications are among the possibilities that will be explored during his presentation.

IBM's ViaVoice enhanced version for Mac is the most viable straightforward option at this time. below are listed some of the features as well as the system requirements.

Key Features:

USB connectivity to ensure high quality digital input with USB microphone included in the box. Direct Dictation with correction into Appleworks 6.0, Microsoft Word 98 and Internet Explorer 4.5. An Enhanced Correction Interface makes correcting dictation faster and easier. Online Command and Control of AppleWorks and Microsoft Word Accuracy improvements including Unsupervised Enrollment, Smart Redictation, and automatic document analysis.
Specialized vocabularies for Computers, Business and Finance, Chatter's Jargon and Cuisine; which can increase accuracy when users work within these disciplines. VoiceCenter user interface with two views (not zoomed and zoomed).

System Requirements:

iMac, iBook, Power Mac(tm) G3 and G4, PowerBook(r) G3 (with factory-installed USB port)
Mac OS 9; version 9.0.4 and higher, but not including MAC OS X
300 MHz PowerPC(r) processor or higher
64 MB available RAM (128 necessary to be productive)
250 MB available hard disk space
CD-ROM or DVD-ROM drive
Audio Output Jack
Adobe Acrobat Reader v.4.x or higher (for viewing PDF documents - included on ViaVoice CD)
QuickTime 4.0

Another option that is available is iListen from Mac Speech and some of its features are listed below.

Feature Summary

Continuous Speech Recognition - Dictation in normal tone of voice and at a normal pace using the Philips FreeSpeech 2000 speech engine
Direct Dictation - Dictate, edit and format text directly in any application
Text Macro Capabilities - Insert (up to 32,000 character) standardized or predetermined phrases and text - with a single spoken phrase
Speech Navigation - Open, close and control applications and navigate the desktop by speech
Large Vocabulary - Active vocabulary of greater than 30,000 words. Backup vocabulary of 300,000 words developed in cooperation with Oxford University Press provides more words
Command and Control - Speech control of the computer through built-in normal language and individually created commands
AppleScript Compatible - Ability to take advantage of Apple's built-in scripting language
Internet Ready - Surf the web with your voice
Multiple User Support - Multiple users create their own speech profiles on the same computer Mac OS 9 or greater
Requires a PowerMacintosh G3 or PowerMacintosh G4 computer with at least 128mb of RAM (iListen does not support the use of upgrade cards at this time)

For the more advanced users a third option would be available which would be running Virtual PC along with Dragon NaturallySpeaking.

Virtual PC for MacOS, Version 4 lets you put a fast, powerful PC inside your G3 or G4 Mac, enabling you to run PC applications, access PC networks, share files with PC-based colleagues, and use PC-only Internet services and applications. Choose from a range of pre-installed pre-configured OS's; Windows(r) 98, Me, 2000, or PC DOS.

System Requirements:

Processor: Any G3/G4 Macintosh
Operating System: Mac OS 8.5, 8.6 or 9.x
Available Hard Drive:1 GB
RAM 64 MB recommended

L&H Dragon NaturallySpeaking(r) Professional Solutions

The L&H(tm) Solution Series speech-enables your PC, providing you with an easy way to quickly increase productivity. Use natural, everyday speech and talk to your computer to automate common tasks, customize and integrate applications, and turn your speech directly into text.

Key Features

Natural dictation into your computer at up to 160 words per minute in memos, letters, reports, and e-mail.
Train the software to recognize your speech in as few as five minutes.
Hands-on tutorial for basic dictation techniques and quick start-up.
Comprehensive, customizable vocabulary of 250,000 standard and business terms, with more than 160,000 active words, enables you to add new terms and create vocabularies for different specialties or fields of interest.
Integrated with Microsoft(r) Word, Corel(r) WordPerfect(r), and Lotus Notes(r); and works with virtually all Windows(r)-based applications
Saves a recording of your dictation session for future playback and revision.
Text-to-speech technology enables your computer to read e-mail and other documents aloud.
Navigation of the Web by voice.
Custom voice commands (macros) let you create documents, insert boilerplate text, customize applications, complete forms, and perform other complex tasks with a spoken word or phrase.

System Requirements

266 MHz Intel(r) Pentium processor with MMX(r) or equivalent;
Windows(r) 95 C, Windows(r) 98, Windows(r) 2000, Windows(r) Millennium, or Windows NT(r) 4.0 (with SP-6 or greater);
128 MB RAM;
195 MB free disk space (150 MB to install Dragon NaturallySpeaking(r) 5.0, 5 MB to install L&H(tm)
SpeechLinks and 40 MB to install L&H(tm) SpeechDocs Filler);
L&H(tm)-Certified Sound Card (Hardware Compatibility list)
Noise-canceling headset microphone (included).


Go to previous article 
Go to next article 
Return to 2002 Table of Contents 
Return to Table of Proceedings


Reprinted with author(s) permission. Author(s) retain copyright.