音声ブラウザご使用の方向け: SKIP NAVI GOTO NAVI

Web Posted on:December 31, 1998


VOICE RECOGNITION SOLUTIONS FOR CHALLENGED SPEAKERS

Neal C. Meskimen - ATS
PACER Center
Minneapolis, MN
(612) 821-9863
E-mail: nmeskimen@pacer.org

Concert Assistive Technologies
North Branch, MN
(651) 674-0100
Email:ConcertAT@aol.com

Program Summary

Now you can successfully adapt Voice Recognition software to permit challenged speakers to be recognized and understood by others for the very first time.

Program

Growing demand and interest has pushed Voice Recognition to new heights of capability. These improvements, combined with the demand for increased functionality, have changed the communication potential for challenged speakers forever. Smarter speech recognition engines have given us faster response and specialized vocabulary words for specific applications with fewer errors. In the near future we will find built-in speech recognition in the majority of systems, permitting users to simply speak verbal commands or dictate without regard to a specific speaker's voice file. The standard decode of recognized sound patterns have also improved with each new upgrade and release. Continuous Speech has significantly improved productivity, functionality and recognition accuracy. However, challenged speakers often require modifications to enjoy the benefits available to most other users.

Through continuous research and application tweaks, a retraining method has been developed that promises high hope for those whose vocalizations do not follow the basic rule of speech in any language. This provides new options for those who continuously struggle to be understood. Now, these individuals can finally optimize their given ability to communicate. This discovery doesn't propose that we have reached the ultimate method to translate all heretofore-meaningless sounds into logical discourse, but it does aspire to enable many individuals to communicate better than they ever have before. These developments provide a first step toward ending the frustration of the inability to interact verbally, and thus provide those with communication difficulties a growing light at the end of the tunnel of constant misunderstanding.

Not all challenged speakers are unable to articulate their thoughts. Some speech-impaired people are quite capable of producing speech on par with their peers in schools, business and society. But they are not always clearly understood. It is to this end that you will hear and observe the remarkable sound conversion process that levels the playing field for so many challenged speakers. The presentation will cover the following:

1. The current consensus of expectations regarding the successful outcome of voice recognition use:

  • A. The user can speak clearly and consistently.
  • B. They can recognize and correct mistakes, most typically caused by inconsistent pronunciation.
  • C. Be able to speak in a somewhat continuos manner.
  • D. Be able to use the keyboard and mouse for selecting and implementing specific program commands.
  • E. Habitually use the program so as to build a strong voice file that has high accuracy.

Item A is the most critical issue in having the speech recognition engine differentiate sounds, which could sound like a quasi-foreign language to most listeners.

Considering items C & D; there are solutions for challenged users depending on the type and severity of the disability. Typically, an involved user would consider using the older programs that required discrete speech with a small pause between each word. Likewise, some of the older programs also provide full options to speak all commands, thus removing the need to use hands and fingers.

Item E is necessary to have on going-success with minimum recognition errors and frustrations.

2. After instruction on the basic consensus of expectations, this presentation will guide the participant through the current state of the art methodology for retraining the voice file soundbase to effect a translation of ostensibly senseless sounds into meaningful communications. This will soon be announced as a product called Clear:Voice (TM), which is patent pending.

The inability to communicate effectively is a frustration that hundreds of thousands of people live with each day. It presents difficulties in almost every aspect of life, and affects the potential to succeed in school, the workplace, family life, and society. This new paradigm of Assistive Technology promises to satisfy a basic human need by providing individuals with the monumental achievement of making themselves understood. Those of us who mastered this skill at two years old can never fully appreciate what a watershed event this is in anyone's life.

Neal C. Meskimen ATS - PACER Center, Minneapolis, MN - December 1998