does robotic sound like voice come from small network or alignment issues ? (in fastspeech2 plus mb-melgan)

TensorSpeech / TensorFlowTTS

:stuck_out_tongue_closed_eyes: TensorFlowTTS: Real-Time State-of-the-art Speech Synthesis for Tensorflow 2 (supported including English, French, Korean, Chinese, German and Easy to adapt for other languages)

https://tensorspeech.github.io/TensorFlowTTS/

Apache License 2.0

3.84k stars 815 forks source link

does robotic sound like voice come from small network or alignment issues ? (in fastspeech2 plus mb-melgan) #284

Closed manmay-nakhashi closed 4 years ago

dathudeptrai commented 4 years ago

i do not know :D. Maybe both :d

manmay-nakhashi commented 4 years ago

@dathudeptrai can it be something other then this two ? :sweat_smile:

dathudeptrai commented 4 years ago

@manmay-nakhashi dataset quality :))). FastSpeech is limitted, i still not better than Autoregressive model as Tacotron2 in all datasets :D.