- 作者: Hwai-Tsu HU, His-Tsung WU
- 中文摘要: This paper presents a glottal excited linear predictive (GELP) speech coder coded at 1.9 Kbps. While the processing of unvoiced speech and silence intervals is performed with a stochastic codebook, a glottal codebook with 32 entries for voiced excitation is used to simulate the glottal-related characteristics of the residual. Detailed procedures are described for extracting glottal features and for generating the excitation. Besides the modeling of phase characteristics, the resulting glottal excitation takes pulse dispersion and turbulent noise into account. Listeners involved in a mean opinion score (MOS) test exhibited strong preference for GELP over the 2.4 Kbps linear prediction coder (LPC-10e), though the overall performance of GELP is somewhat inferior to that of the 4.8 Kbps code excited linear prediction (CELP) coder.
- 英文摘要: --
- 中文關鍵字: speech coding, glottal excitation, linear prediction, glottal phase characteristics, glottal closure instant
- 英文關鍵字: --