An Adaptive Approach to a 2.4 kb/s LPC Speech Coding System.
Abstract
The goal of this research was to investigate (a) Adaptive estimation methods for noise suppression and performance enhancement of Narrowband Coding Systems for speech signals and (b) Autoregressive spectral estimation in noisy signals for speech analysis applications. Various prefiltering techniques for improving linear predictive coding systems were investigated. Filter coefficients were varied to optimize each filter technique to remove noise from the speech signal. A new prefilter consisting of an adaptive digital predictor (ADP) with pitch period delay was developed and evaluated. The one-dimension filter approach was expanded upon to use a two-dimensional approach to suppress noise in the short time fourier transform domain. The two dimensional approach was found to have significant potential. Fast algorithms for efficient solution of the linear estimation problem and a new recursive linear estimator suitable for rapid estimation of a signal in noise were developed. Keywords include: Speech analysis; Pitch estimation; Noise suppression; Signal estimation; Adaptive digital filter; and Speech enhancement.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jul 01, 1985
- Accession Number
- ADA160312
Entities
People
- C. S. Sims
- D. L. Soldan
- L. D. Hoy
- R. D. Preuss
- R. Yarlagadda
Organizations
- Oklahoma State University–Stillwater