Automatic Speech Recognition (ASR) technology is the independent, computer-driven transcription of spoken language into readable text (Stuckless, 1994). The technology commonly has been used to prevent or alleviate workplace cumulative trauma injuries such as carpal tunnel syndrome, and to enable computer access for individuals with fine motor limitations, vision impairment, or learning disabilities.

JAN consultants often are asked about using ASR technology as an accommodation for individuals who are deaf or hard of hearing in situations involving group meetings, or telephone communication. While ASR technology has applications in some situations, such as in interviews or one-on-one meetings with supervisors, the technology still has limitations in groups. ASR systems, such as the Dragon family of speech recognition products and IBM's ViaVoice, are single-speaker systems as opposed to speaker-independent systems. That is, the software must be "trained" to recognize each user's commands. Because ASR systems only recognize speakers who have trained the software, it may not be effective when used in situations where multiple speakers must be recognized at one time or for conversations over the telephone with various people.

One approach to using ASR with multiple speakers is "voicewriting." This technique involves having one individual who has trained the ASR repeat the information spoken by others. The information repeated by this person then is translated by the ASR into real-time text. Voicewriting can be an effective accommodation for staff meetings, conferences, or training situations. For more information about voicewriting services and technology, see JAN's Web site, at http://www.jan.wvu.edu/cgi-win/OrgQuery.exe?Sol555

Telephone uses for assisted ASR also are under development. Ultratec, Inc. has developed, and is consumer testing, the CapTel phone which automatically connects to a captionist who uses speech recognition software to transcribe the call into text. The communicating parties do not hear the captionist. The party using the CapTel phone reads the transcribed text in real-time. Although the CapTel phone is not yet available to the general public, it is likely that such telephone captioning services will be used in the near future as a job accommodation for individuals who are deaf or hard of hearing.

Lastly, advances in speech recognition technology have lead to the development of products specifically designed to improve communication access for people who are deaf or hard of hearing. Technology such as the iCommunicator and the Signtel Interpreter use speech recognition software to facilitate personal communication via text, video sign language, and computer-generated voice. These applications are being used in the workplace to facilitate effective communication.

For more information about accommodations for people who are deaf or hard of hearing, go to: Work-Site Accommodation Ideas for Individuals who are Deaf or Hard of Hearing at http://www.jan.wvu.edu/media/Hearing.html.

Resources:

Stuckless, R. (1994). Developments in real-time speech-to-text communication for people with impaired hearing. In Stuckless, R., Special Applications of Automatic Speech Recognition (ASR) with Deaf and Hard-of-Hearing People. Retrieved from the Internet, March 13, 2003 at http://www.netac.rit.edu/publication/asr.html.

- Tracie D. Saab, MS, Human Factors Consultant