Speech Recognition

Contributed by Christine Oddo (UB/CAT)

Features Eligible Students Student Abilities Aids & Supports Resources

A speech recognition system is another tool for writing. It provides an alternative option for students unable to use standard methods for computer access to write more efficiently. It should be viewed as part of a continuum of writing technology strategies that includes pencils to computer input by voice.

With speech recognition software and computer hardware, a student trains the computer to recognize her voice for writing or giving computer commands. Students can use voice recognition to write within a word processor or to create an email message. It can also execute program commands such as saving or printing or to control the mouse.

Portable Use

Some programs can be used with portable digital recorders to let students record their speech in multiple environments. The computer then transcribes their work to text later.

Features

Speech recognition is a rapidly growing and changing technology field. Many companies offer more than one version of a program; different versions will have different features. As you investigate these programs look for features that are most beneficial for individual students.

  • Continuous Speech -- A student speaks a string of words with a normal rate of speech, not pausing between words.
  • Dedicated Word Processor -- A built-in word processor provides the user a quick start.
  • Hands-Free -- A student's voice can create text, input commands and perform mouse movements. Other programs require using the keyboard/mouse together with the student's speech.
  • Playback - The spoken text is repeated in the student's voice.
  • Text-to-Speech - The produced text is read back with a computer voice. This may make it easier to identify errors.
  • Use Across Applications - Some programs work with one or two applications, such as word processors. Others work with all software programs.
  • Customization - A variety of options should be examined:
    • Built-in keyboard commands
    • Built-in voice navigation
    • Built-in macros
    • Macros created by the user

    Macros provide opportunities for students to control repetitive tasks, insert pre-determined text phrases, or create scripts to do things like send e-mail, select and complete math problems or organize a personal calendar.

Several programs are available that provide a range of options. This promising technology will most likely be the input of choice by a wide range of computer users. Check out the ATTO Speech Recognition Software Comparison Chart for additional information on features appealing to students.

Programs that offer continuous speech include:

Products using Dragon Naturally Speaking are provided by ScanSoft, Inc. Check their Product Feature Matrix on the website to compare models.

Discrete Speech - With discrete speech, a student must pause between each spoken word. This feature was common with the first speech recognition programs. Then it was limited for a while, but today, the Naturally Speaking products again allow for "one word at a time" speech. The slow pacing and operation may best match the student's production style, at least at a certain point in time. Custom words can be entered into the active vocabulary. This is useful for students whose voices are not strong or consistent. Programs offering discrete speech include:

  • Dragon Dictate (Dragon Systems, Inc.)
  • Keystone Screenspeaker (KeySpell)
    This training utility for use with Dragon voice recognition software helps poor readers set up and use their dictation software. It reads your work back to you as you dictate.

 

Back to top
Eligible Students

Students who can benefit from speech recognition vary greatly in their needs. Speech recognition can assist students who have the following difficulties:

  • Students who are unable to manipulate writing tools and who require an alternative system.
  • Students who have difficulty with writing mechanics; those having difficulty with spelling, grammar, etc.
  • Students who have difficulty with the writing process; with getting their ideas written down.

Back to top

Student Abilities

Successful use of speech recognition requires certain abilities of student users. The level of independence will vary with each student. Students must:

  • Be able to complete the voice training program,
  • Be able to understand the dictation process,
  • Learn how to use the voice program; be able to repeat phrases in the same level of intonation
  • Remember verbal prompts
  • Identify the correct spelling of words,
  • Identify words as they appear on the screen and recognize and correct as they occur, and
  • Be able to learn how to use the speech program with other applications.

View our voice recognition checklist for additional information.

 

Aids & Supports

Students using speech recognition will need support from the educational system in terms of:

  • Sufficient training for the student with trainer available,
  • Technical support from persons familiar with the program,
  • Adequate release time for training of technical persons, educators, family members,
  • Quiet school locations where the students can use the system satisfactorily, and
  • Instructional support for the student to progress with writing opportunities.

Back to top
Voice Recognition Resources

Within this fast changing field, there are excellent resources to keep you updated:

From the QIAT listserv, here are Tips for Using Speech Recognition with Students with Disabilities.

Speaking to Write
An excellent site for issues related to speech recognition technology. Join the spk2wrt listserv, explore the message archives or interact with a variety of materials on this topic.

Say I Can!
Read the entire Dragon Naturally Speaking Guide free! All twenty-one chapters (288 pages) of this top-selling book are yours to read online.

Speaking To Write/ Word For Word: An overview of Speech Recognition by Bob Follansbee, Ed. D.
Find out what we really know about speech recognition for students with LD and how it works or doesn't.

NCIP Voice Recognition
This site offers updates on voice recognition technology with input from Dr. Bob Follansbee, Director of the Computer Learning Program at Children's Hospital in Boston, Massachusetts.

Back to top

copyright © 2000 - 2005 Assistive Technology Training Online Project