Skip to main content

Table 4 Number of free parameters of adaptation techniques

From: Speaker adaptation in the maximum a posteriori framework based on the probabilistic 2-mode analysis of training models

Method Number of free parameters
Probabilistic 2-mode based model 20·35 (K R ·K D )
Tucker 3-mode based model 20·35 (K R ·K D )
MLLR, 3-block-diagonal regression matrix 13·40
MLLR, full regression matrix 39·40
MAPLR adaptation 39·40
Eigenvoice 50