Speech Sources Subscribe SpeechTEK Industry Jobs
• Current Issue
• STM NewsBlast
• Industry News
• Industry Links
• Industry Whitepapers
• Industry Casestudies
• Magazine
• Back Issues
• STM NewsBlast
• Conference
• Exhibitor
• Registration
• Ad Info
• Contact Us
• Editorial Calendar
• Submissions
• Post a job
Advanced Search

See Current Issue

April 13, 2004

Send to a Friend Printer-Friendly Version
AVIOS~SpeechTEK 2004 Awards

Leon Lerman Best of Show:

Awarded to Unveil Technologies Inc. for its "Conversation Manager," a contact center speech application that dynamically learns and improves customer experience.


Best Telephony Product:
Awarded to Brooktrout Technology for its new TR1000™, a telephony voice platform that offers a broad range of media processing products all under a single API.


Best Development Environment:

(Two winners)

Awarded to Nuance for its Application Environment (NAE), a fully integrated application design, development and debugging environment.


Awarded to Intervoice for InVision Studio, a fully VoiceXML compliant graphical development environment that streamlines and enhances the creation of voice solutions.


Best Application Management Product:

Awarded to VoiceObjects for its VoiceObjects Factory™, an open standard voice application management system for fast and easy creation of voice-controlled services.


Best Data Mining Application:

Awarded to Nexidia for its intelligent mining application, a phonetics-based solution that can search more than 30 hours of audio in less than one second.


Best Voice Portal Approach:

Awarded to Angel.com for their low-cost custom voice solutions (Voice Sites) that leverage the Web to build, manage and deploy applications for the SME market.


Best Self-Service Application:

Awarded to NetByTel for its Whirlpool Service Scheduler, a voice application that has been built as an extension of Whirlpool’s call center application, leveraging the existing scheduling service engine.


Best New Speech Technology:

Awarded to BBN for successful large-scale commercial deployment of semantic interpretation of customer speech.


Best Packaged Application:

Awarded to ScanSoft for its SpeechPAK 1.0, the first application kit of its kind to be offered to the healthcare industry, designed to help support member and physician self-service initiatives.


Gary Poock Best Paper Award:

David T. Williamson and Mark H. Draper, from the Air Force Research Laboratory and Timothy P. Barry from Sytronics Inc., authors of "Commercial Speech Recognition Technology in the Military Domain: Results of Two Recent Research Efforts."


Honorable Mention in the Gary Poock Best Paper Award Category:

Lizanne Kaiser and Sondra Ahlén of Edify Corporation, authors of "Are you Listening to your Spanish Speakers? How Spanish speakers in the U.S. say U.S. addresses."


The Best Paper Awards:

Since 1997, AVIOS has presented the Gary Poock Best Paper Award. All

authors who submit papers for presentation at the annual conference are

considered for the award.


Submitted papers were judged on the clarity of the writing, the soundness of

the methodology, depth and completeness of the work, the degree of

innovation represented in the work, value-added to the existing knowledge

base and the potential impact of the work on the industry.


David T. Williamson has a long history of presenting research at AVIOS conferences on creative military applications of speech recognition technology. In this year's paper, he presented compelling data from two studies showing that speech recognition outperformed keyboard entry in terms of both accuracy and speed in two real-world military tasks and has now reached an accuracy level acceptable to the military for operational deployments.


Lizanne Kaiser and Sondra Ahlén explored some issues of users whose first language is Spanish. In speech applications requiring saying United States addresses, native Spanish speakers often mix characteristics of both languages in their speech (e.g., using the word order of English, but the words of Spanish). Word pronunciations also vary widely, not always following the standard rules of Spanish pronunciation but rather the conversational characteristics of this user population. Developing speech applications for diverse populations requires deeper understanding of the conversational characteristics of particular user populations.


Both papers were excellent examples of how careful data analysis can lead to lessons of importance for those who design and deploy speech applications.