An Algorithm for Extraction of Speech Spectral Envelope Using Piecewise Linear Predictive Coding
Wan Mao-wen
Abstract
Wan Mao-wen
Abstract
In speech analysis, it′s very important for the extraction of speech spectral envelope. An algorithm is presented using piecewise LPC(Linear Predictive Coding). The theory of filter banks is used as well as LPC. The spectral coverage of speech is divided, and then LPC is applied to different spectral bands. Experiments suggest that the resolving capability of synthetic spectra is improved.
A significance statement is not available in the OpenAlex record.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
In speech analysis, it′s very important for the extraction of speech spectral envelope. An algorithm is presented using piecewise LPC(Linear Predictive Coding). The theory of filter banks is used as well as LPC. The spectral coverage of speech is divided, and then LPC is applied to different spectral bands. Experiments suggest that the resolving capability of synthetic spectra is improved.
Key concepts: Linear predictive coding, Spectral envelope, Speech coding, Predictive coding, Linear prediction, Speech recognition, Envelope (radar), Coding (social sciences)