StarGAN-Voice-Conversion icon indicating copy to clipboard operation
StarGAN-Voice-Conversion copied to clipboard

Published 20 hours ago • verified publisher icon hujinsen

not able to reproduce the results in original paper

Open NOrangeeroli opened this issue 4 years ago • 6 comments

I cannot reproduce the results in original paper. I used tensorflow 1.8.0.

NOrangeeroli avatar Mar 17 '20 04:03 NOrangeeroli

I also used tensorflow 1.13.0 and the result is not similar to the original paper

malradhi avatar Apr 13 '20 20:04 malradhi

@malradhi How about the wav converted? My generated wavs just like noise

Liujingxiu23 avatar Apr 14 '20 07:04 Liujingxiu23

@Liujingxiu23 My generated waves are not understandable at all. Did you find a model which produce a good non-parallel VC so far?

malradhi avatar Apr 14 '20 21:04 malradhi

@Liujingxiu23 My generated waves are not understandable at all. Did you find a model which produce a good non-parallel VC so far?

Did u tried StarGAN2-vc? I searched on git however there are no mature implements.

Lukelluke avatar May 02 '20 10:05 Lukelluke

@Liujingxiu23 My generated waves are not understandable at all. Did you find a model which produce a good non-parallel VC so far?

Did u tried StarGAN2-vc? I searched on git however there are no mature implements.

Thank you, but I can't find the github for StarGAN2-vc.

malradhi avatar May 10 '20 01:05 malradhi

@Liujingxiu23 My generated waves are not understandable at all. Did you find a model which produce a good non-parallel VC so far?

Did u tried StarGAN2-vc? I searched on git however there are no mature implements.

Thank you, but I can't find the github for StarGAN2-vc.

I did find some implements about StarGAN-VC2 via pytorch on github, however still can't generate better covert results . What a pity .

Lukelluke avatar May 17 '20 09:05 Lukelluke