|
|
|
Keynote Speaker Paul Bamberg - Dragon Systems "Digital Sound" Ever since the introduction of CD-ROM technology, the digital capture and broadcast of sound has captivated the imagination of engineers, scientists and artists. Voice activation, speech recognition and digital music are some of the key research areas in the current topics in computer science. We discuss various aspects of sound technology with particular emphasis towards speech recognition. The second part of the talk is a tutorial on how to write your own digital recorder under Windows. One of the least documented features of the Windows operating system is the low-level audio interface. This talk explains how sounds are played back, recorded, and saved to disk in Windows and demonstrates how to create a sound-recording program for a personal computer that, unlike the Windows Sound Recorder, automatically detects the start and end of speech and trims off excess silence. Paul Bamberg, a graduate of Harvard and Oxford, was a member of the Harvard faculty from 1967 to 1995. Thousands of physicians have learned most of what they know about physics from his courses for premedical students in Harvard College, in the Extension School, and in the Summer School. As Director of Science Instruction for Continuing Education, Paul built up a substantial computer science program for the Summer School and Extension School from scratch, frequently by signing himself up to teach whatever key courses could not otherwise be staffed. In 1982 Paul helped to found Dragon Systems, currently the market leader in commercial large-vocabulary speech recognition. Now retired form (*from*) Harvard in order to be a full-time Vice President and Dragon Fellow, he has done research and product development in all key areas of speech-recognition technology -- signal processing, acoustic modeling, language modeling, and natural-language processing.
|
|
|
|||||
|
© 1999 President and Fellows of Harvard College