sprocket
sprocket copied to clipboard
k2kobayashi
Voice Conversion Tool Kit
Which values in the configuration files do I need to adjust to get the highest quality output? I increased n_iter and n_mix substantially, but the output sounds the same as...
Dear Mr. Kobayashi and Mr. Toda, I found following issue while rewriting sprocket to C++: In feature extractor, hop size is given in milliseconds, while in synthesizer, it is given...
Dear Maintainer, I have been trying out sprocket for some standard voice conversion tasks. Although no thorough benchmark has been done, it came to my notice that sprocket trains a...
I was able to run example upto initialize configure files for the source and target speakers and the speaker-pair. But when i try to trace f0, i,e python initialize.py -3...
The project is great and it tunes perfectly however, the pitch is getting too high too sound like mickey mouse. Is there any way to produce more human like and...
Hi, First, thanks a lot for sharing the code! In VCC2018, the sprocket is used also for the non-parallel task. Are there any guidelines / examples for using the sprocket...
Hi, thanks for wonderful project I'm using your project and after a few samples, I realize that the output duration (including pause, speed, etc) always more likely to be similar...
Thanks to #111, download_vcc2016data.py is not necessary. its removal will be excused end of this month.
話者名とwavディレクトリが一致しない場合にファイル名に対応したディレクトリを作成できずにエラーで落ちる． どちらかに統一する必要がある． listに書かれたディレクトリを，当該ディレクトリ（例えば，h5_dir）以下にネストを保持しつつ，作成する方が無難と思われる．
### ToDo - [ ] Upload conversion samples - [ ] Create document and open at GitHub.io? - [ ] Register PyPI - [x] Test using VCC 2018 dataset -...