Speech quality and speaker characteristics
J.H. Eggen, S.G. Nooteboom
Abstract
J.H. Eggen, S.G. Nooteboom
Abstract
One of the applications of speech-coding algotithms is the generation of the output speech of text -to-speech systems. Linear predictive coding (LPC) is still one of the most powerful algorithms used for this purpose. This chapter describes some experiments which evaluate the quality of LPC speech.In particular, intelligibility and naturalness of LPC speech are studied. Ways are indicated to improve existing LPC speech-coding schemes. One altemative, the glottal-excited LPC synthesizer, is described in more detail.
OpenAlex reports 1 citations for this work. Citation counts describe recorded attention and do not establish research quality.
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.
One of the applications of speech-coding algotithms is the generation of the output speech of text -to-speech systems. Linear predictive coding (LPC) is still one of the most powerful algorithms used for this purpose. This chapter describes some experiments which evaluate the quality of LPC speech.In particular, intelligibility and naturalness of LPC speech are studied. Ways are indicated to improve existing LPC speech-coding schemes. One altemative, the glottal-excited LPC synthesizer, is described in more detail.
Key concepts: Linear predictive coding, Speech recognition, Speech coding, Intelligibility (philosophy), Computer science, Naturalness, Voice activity detection, Codec2