Application of Shannon's Rate Distortion Theory to Speech Encoding,
Abstract
The digital encoding of speech is considered from the point of view of Shannon's rate distortion theory. First, a non-adaptive predictor-subtractor operating at intervals less than the pitch period is used to whiten the speech signal. This makes possible the derivation of upper and lower bounds on Shannon's rate distortion function which gives the minimum rate necessary to achieve a specified mean square error. Finally, a subjective test was performed to determine the relationship between speech quality and the value of the rate distortion function. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Sep 15, 1972
- Accession Number
- AD0750537
Entities
People
- Huseyin Abut
Organizations
- North Carolina State University