Module 4 - Assistive Technology for Computer Access (P.6 of 10)
|
|||
Alternative Text Entry - Speech Recognition Software Providing an individual with a disability with an alternative keyboard as a computer input option is not always sufficient. Due to the individual's disability, he or she may not be able to access any of the types of alternative keyboards that have been discussed. Others may find alternative keyboard usage too slow and desire a faster typing option. An alternative text entry option can provide solutions to both of these issues. The most popular alternative text entry option is speech recognition software which provides hands-free computer access and allows users to enter up to 250 words per minute. What is Speech Recognition Software? Speech recognition software is a computer access option that allows the user to dictate text into a computer or navigate the computer using his/her voice to produce words or dedicated voice commands. Speech recognition software works with most standard Windows-based computer programs including word processing programs, email programs, spreadsheet programs, and the internet among others. Many speech recognition programs also provide options for controlling the mouse via voice commands. Navigating the computer, controlling the mouse, or editing all involve remembering specific commands and thus can add an increased cognitive load to using the program. Who Benefits from Speech Recognition Software? Speech recognition can obviously help people with physical disabilities who cannot use a keyboard to access a computer, but it can also benefit people with visual or cognitive disabilities. People with visual impairments can benefit from speech recognition software to enter text if their visual impairment causes difficulty in seeing keyboard keys and thus typing. People with cognitive disabilities often have difficulty spelling, thus their writing suffers. Speech recognition software does not spell words incorrectly, alleviating this problem and allowing them to focus on writing content. It does on occasion enter the incorrect word, however there is an option in place to help recognize when this occurs. Speech recognition programs have a read-back function that reads what has been entered so that people with visual or cognitive impairments have an audio option for checking their work. Due to its ease of use and hands-free nature, speech recognition software is a beneficial computer input option for people with various disabilities. How does Speech Recognition Software Work? Most speech recognition software is speaker dependent, so each user must create their own voice profile. From then on, the user must activate that profile to use this software effectively. Initial voice profile creation also entails training the computer to recognize the user's voice. To do this, the user participates in some program-designated training involving reading various words or phrases so that the computer can "learn" the user's voice. Accent, inflection and tone all effect how the computer recognizes words and this initial training allows the computer to create a baseline for each of these aspects of the user's voice. With each use the program becomes more familiar with how the user speaks. This information is saved in the voice profile which is constantly updating and thus improving recognition. To help the computer improve accuracy, it is important to make corrections when the program inaccurately recognizes words. These corrections are noted in the user voice profile and help to improve future recognition. Although most speech recognition programs claim 98%+ accuracy from the beginning, the reality is that initially accuracy is anywhere from 75%-98%. The more a user trains and uses the program, the better the accuracy becomes. Even if the user has a speech impairment, the program can become very accurate if the user is willing to put in significant training time. What Tools are Needed to Use Speech Recognition? Speech recognition software requires the use of a microphone. The most commonly used type is a headset microphone. This type of microphone can be positioned close to the user's mouth, moves with the user's head, and tends to block out background noise. Thus it provides the most accurate recognition. However, headset microphones are not always the best option. If the user has a disability that prevents the use of his/her hands, it can be difficult or impossible to physically put on a headset. In this case, desktop microphones, which remain stationary, are commonly used. Although desktop microphones have the advantage of being hands-free, they do not move with the user, thus slight head movement can result in varied accuracy. These microphones also have a tendency to pick up more background noise than headset microphones, which can effect accuracy. Regardless of the type of microphone used, for consistent results, the user should always use the same microphone. What is the Availability of Speech Recognition Programs? Speech recognition software was developed for the able-bodied population as an alternative means of text input. For people with disabilities, this means that this product is readily available and at affordable prices. Speech recognition software is available for both Windows-based and Mac computers, however it is much more advanced in a Windows setting. Windows has speech recognition software built in for dictation purposes only in their XP systems and for total computer control in their Vista systems. The preeminent speech recognition program, Dragon Naturally Speaking, is only available in a Windows format. Mac users have a couple of options for speech recognition software in iListen and ViaVoice, however, ViaVoice has suspended research and development and iListen is inconsistent at best. The best option for a Mac user may be to run Windows on their Mac and then access a supported program such as Dragon Naturally Speaking. |
|