Speech Recognition: From Laboratory to Consumer - Half a Million Callers a Day can't be Wrong!

Hy Murveit, Nuance Communications
hy@nuance.com

Seminar on People, Computers, and Design
Stanford University January 8, 1999

 

This talk will survey the transition of speech recognition technology from laboratory to telephone call-center applications that you may have used. We will survey the laboratory sources of the technology (state of the art, trends and challenges). We will then summarize the process of moving technology from the laboratory into a growing number of telephone-based systems. Such systems are fast becoming an important way for businesses to provide important information to their customers and employees. These systems offer cost savings to the deployer, convenience to the consumer, and they are as available as the nearest telephone (which might be in your pocket). Charles Schwab (stock quotes), UPS (package tracking), American Airlines (AAdvantage numbers), and Sears (call routing) all deploy such Nuance Communications systems. These and other companies using Nuance systems currently automate 500,000 calls per day. We will describe some of the challenges in these applications and some of the lessons learned.

Hy Murveit is Vice President of Research and Development at Nuance Communications, a spin-off of SRI International that develops speech recognition, natural language and speaker verification software to automate access to information and services over-the-phone. Dr. Murveit was one of the four founders of Nuance Communications. Dr. Murveit previously worked at SRI's STAR laboratory from its inception in 1983 where he was an algorithm developer and principal investigator for government-funded research projects. Hy received his Ph.D. from U.C. Berkeley in 1983, also focusing in the area of speech recognition.

 

Titles and abstracts for all years are available by year and by speaker.

For more information about HCI at Stanford see

     h