Multisensor Speech Input.
Abstract
The use of multiple sensors to transduce speech was investigated. A data base of speech and noise was collected from a number of transducers located on and around the head of the speaker. The transducers included pressure, first order gradient, second order gradient microphones and an accelerometer. The effort analyzed this data and evaluated the performance of a multiple sensor configuration. The conclusion was: multiple transducer configurations can provide a signal containing more useable speech information than that provided by a microphone. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Dec 01, 1983
- Accession Number
- ADA140894
Entities
People
- K. F. Karnofsky
- K. N. Stevens
- M. N. Alakel
- V. R. Viswanathan
Organizations
- BBN Technologies