Skip to main content
EURASIP Journal on Audio, Speech, and Music Processing

Table 1 Summary of the streams used in the HMM-based synthesis system

From: A uniform phase representation for the harmonic model in speech synthesis applications

 

Stream 1

Stream 2

Stream 3

Harmonic Model + Phase

   

Distortion (HMPD)

   

Content

f 0

Amp.Env.

σ i (f)

Parameters

log

MCEP (39)

MCEP (12)

Model

Cont.HMM

Cont.HMM

Cont.HMM

STRAIGHT [10],[28]

   

Content

f 0

Amp.Env.

Aperiodicity

Parameters

log

MCEP(39)

Bandwise (5)

Model

MSD-HMM

Cont.HMM

Cont.HMM

  1. Amp.Env., amplitude envelope; MCEP(39), Mel-CEPstral coefficients of order 39; Cont.HMM, continuous HMM model; MSD-HMM, Multi-Space Distribution (MSD) [32].
Back to article page