Desription:
  GMM-MCEP-GV: Considering GV at parameter generation in GMM-based conversion with mel-cepstra;
  GMM-MCEP-GV: GMM-MCEP: GMM-based conversion with mel-cepstra as the spectral parameter;
  EL: Electro-Larynx speech;
  HE: HElthy speech;
  The values in the table are the mel cepstrum distortion (without power information, speech frames only, witout global variance) in dB; Click on the value for a speech sample
Wav samples: Matched Case
Acknowledgement: