Title :
Noise compensation for linear prediction via orthogonal transformation
Author_Institution :
Dept. of Electron. Eng., Nat. I-Lan Inst. of Agric. & Technol., Taiwan
fDate :
8/1/1996 12:00:00 AM
Abstract :
An efficient method is presented for compensating for the bandwidth broadening effect while performing linear prediction of noise-contaminated speech signals. This method takes advantage of an orthogonal transformation in vector space. The computational load is negligible, but the improvement as manifested by the sharpened formant peaks is considerable
Keywords :
compensation; prediction theory; spectral analysis; speech processing; bandwidth broadening effect; computational load; linear prediction; noise compensation; noise-contaminated speech signals; orthogonal transformation; sharpened formant peaks; vector space;
Journal_Title :
Electronics Letters
DOI :
10.1049/el:19960951
