|
Volume 16, No. 2, Spring 2008 |
Subscribe to AT Messenger Download PDF Viewer |
| PDF Version (for printing) Large Print (PDF) Text Version |
Chris Pennington, MS
Research Coordinator
AgoraNet, Inc.
James Polikoff, MS
Senior Research Associate
Speech Research Laboratory Alfred I. duPont Hospital for Children
The ModelTalker Speech Synthesis System is a revolutionary software package designed to benefit people who are losing or who have already lost their ability to speak. It allows people with ALS (Amyotrophic Lateral Sclerosis) or other conditions to create a synthetic version of their own voice for communication, or to choose a voice best suited to represent them. The system uses representative segments of recorded speech (either from an individual who is losing the ability to speak or from a person chosen by someone who can no longer speak) to create a unique personalized synthetic voice. This synthetic voice is virtually unlimited, meaning it can be used to express almost anything, including words and phrases not previously recorded. The voice “banking” process is guided by computer-assisted voice recording software called InvTool. Once this process is finished, an individual will be able to create a synthetic voice by uploading their recordings to a voice generation site. The voice, when combined with the ModelTalker speech engine, can then be used with any communication system that is SAPI 5.0 compatible.
The ModelTalker project is a collaborative effort between AgoraNet, Inc. and the Nemours Speech Research Laboratory. AgoraNet, Inc. is a custom software and web development company located in Newark, Delaware. Established in 1996, they have been significantly involved with numerous assistive technology and multimedia development projects. In addition, their staff has expertise in the areas of augmentative communication, human-computer interfaces, natural language processing, and accessibility issues. The Speech Research Laboratory is one of the laboratories within the Nemours Center for Pediatric Auditory and Speech Sciences (CPASS) at the Alfred I. duPont Hospital for Children in Wilmington, Delaware. The Speech Research Laboratory conducts research to develop, evaluate and refine speech analysis, synthesis, and recognition technology for children and others with speech and hearing disorders.
Currently, the complete ModelTalker System is in the early stages of commercial development and the developers are actively seeking beta testers both in and outside of the ALS and AAC communities—anyone can be a beta tester. Beta testing software is a great way to experience and help shape the development of new software programs. It often gives you access to tools you may need months before they are available to the general public, and it gives you a way to influence the development of that software by having direct access to the developers during the creation process. With this opportunity, however, comes some responsibilities. First, you must remember that this software is under development. The software is stable, but there is the possibility that not all functionality is intact and that it may even crash. The developers don't think so, but that is where they need your input, ideas, and the diversity of your experience and equipment. Second, you must realize that there is no telephone technical support for these products. Once the products are commercially released, of course, full support services will become available. If you would like to be a beta tester, the software can be downloaded for free from www.modeltalker.com. There you can learn more about a short inventory of about 14 words and phrases for review. The InvTool program and accompanying tutorial guide you through the entire process. Once the test speech files are uploaded, personnel at the speech lab will look them over and possibly make additional suggestions for creating better recordings. If all is well, the lab will send you a full inventory of 1650 words and phrases to record. You should expect that recording the full inventory will take at least 8 to 10 hours distributed over 3 or 4 days; for some people it can take a lot longer. When all of the phrases are recorded, you will upload the speech files to the speech lab for conversion to a synthetic voice. You will receive a web link to download your voice once it has been created.
For more information, please visit www.modeltalker.com.
