I learned how to type when I was in high school. I remember sitting in the classroom wondering why I needed to learn this “important” skill. But the truth is, as I grew older and my educational challenges increased, I found myself pounding on the keyboard more and more.
To this day, despite the best efforts of that long-forgotten teacher in high school, I often make spelling errors. And, I still hunt and peck on the keyboard. Spell checkers have helped the first problem. Voice recognition technology has virtually eliminated the second.
Voice Recognition Technology: How it Works
Voice recognition technology converts spoken words into text or computer control commands. There are many people for whom it is a boon. Typing-challenged people like me and those who have upper-extremity impairments all benefit. But perhaps the biggest group to benefit from this technology is people who experience communication disorders.
Because voice recognition technology incorporates a "training" phase, people who have hearing deficits, as well as those who stutter, have dyslexia or aphasia, or mumble can be helped. While the obvious objective of training is to instruct the computer to understand the user (and this is an important part of the process), at the same time the user is getting trained—practicing his or her speaking and getting immediate feedback from an impartial judge. The more attention given to how words and phrases are pronounced, the better they will be interpreted by the software. Ultimately the ability to communicate will also improve.
People with aphasia have difficulty recalling words and formulating sentences. The National Institutes of Health, MossRehab and Unisys Corporation have joined together to help. They’re developing a language therapy system that uses computers equipped with audio-visual feedback and voice recognition software. Through the project, people with aphasia are rebuilding their language skills.
Voice Recognition Technology for Deaf and Hearing Impaired
Another group based in Hungary, has developed a software package that helps to develop speech and reading skills by providing users with automated speech recognition and visual feedback. The software is aimed specifically at those with hearing impairments, for whom auditory cues are virtually useless, but a side benefit is the development of memory and concentration skills.
Communication Disorders Helped
According to Rachel Bress, a speech and language pathologist in Pennsylvania, “People who stutter are always trying hard to control their speech whether it is by slowing down, taking breaths before each word, compensating by using other words, etc.”
One possible solution is a voice recognition program that provides the user with delayed auditory feedback. Users speak into a microphone, and a fraction of a second later hears their voice. Short delays can reduce stuttering by as much as 70 percent, while longer delays produce greater fluency. One company, Artefactsoft, has developed two voice recognition technology programs, one for handheld devices and another for desktop/laptop computers.
Learning Software
Finally, Rupal Patel and Deb Roy, from the University of Toronto’s Department of Speech-Language Pathology, believe the future is in teachable software. The researchers are looking into software that can learn from, and adapt to, the abilities of the individual user. The program would be able to recognize a user’s speech patterns—how words are spoken instead of just the words themselves—thereby giving another piece of information the computer can use to better interpret what is being vocalized.
Training voice recognition technology isn’t easy, especially for someone with speech difficulties. It takes time and effort. Words and phrases have to be repeated. Mistakes have to be diligently corrected. The frustration level of the user can be high. But the system is patient and private.
