From: Voice conversion using speaker-dependent conditional restricted Boltzmann machine
SD-RBM
NN
RNN
GMM
p
0.4417
0.0913
0.3299
0.0000