|
| |
|
alternative input devices
|
|
keystroke modification and mouse emulation
typically provide combinations of one or more of the following: execution of multiple-keystroke commands (such as control-x) serially rather than simultaneously; manipulation of the mouse via the keyboard; alternatives to clicking and/or holding mouse buttons; and visible and/or audible feedback for keystrokes. Both the Macintosh and Windows operating systems provide some of these capabilities.
|
|
joystick and switch
combinations enable an individual lacking sufficient mobility to use a full keyboard to access a computer through an on-screen keyboard or other "virtual" substitute.
|
keyboard alternatives
allow individuals unable to use a standard keyboard to input keystrokes with a mouse, headmount, or other specialized device. Simplified key arrangements and one-handed keyboards are common, but there are also systems which allow input of morse code or with the use of a pointer, such as the Magicwand Keyboard shown here.
|
|
rate enhancement
programs allow users to define shortcuts or macros for frequently typed words or executed commands. The software detects these sequences and expands them in-place, or executes the commands. Also, word prediction programs utilize weighted lexicons to offer automatic completions for started words. Microsoft's Word is an example of a program which does word prediction.
|
|
voice recognition
has improved a great deal over the past several years. Early systems were inaccurate, required training for individual use, and were expensive. However, a continuous-speech, user-independent, large-vocabulary system has yet to be achieved. Since even an individual does not prounounce a word the same way every time, the program must make a best guess. The larger the vocabulary, the more prone the system is to errors. User-independent systems don't have large vocabularies to improve accuracy. And most systems still rely on discrete speech, or clearly distinct words, to achieve acceptable performance.
There are two general categories of voice recognition software. The first is activation/navigation. The user gives simple commands to the program, such as "open file," which the program interprets and translates into instructions for the operating system. The second category is dictation systems. These are moderately to prohibitively expensive programs which are trained to recognize individual users' speech patterns. These programs require a training period before it is possible to use them, but also continue to learn and adapt with each subsequet use. Although 60,000 word vocabularies may seem impressive, keep in mind that the average document includes only a few thousand unique words.
|
|
assistive technology
barriers to access