eduzhai > Applied Sciences > Engineering >

Unsupervised Cross-Domain Singing Voice Conversion

  • king
  • (0) Download
  • 20210506
  • Save

... pages left unread,continue reading

Document pages: 5 pages

Abstract: We present a wav-to-wav generative model for the task of singing voiceconversion from any identity. Our method utilizes both an acoustic model,trained for the task of automatic speech recognition, together with melodyextracted features to drive a waveform-based generator. The proposed generativearchitecture is invariant to the speaker s identity and can be trained togenerate target singers from unlabeled training data, using either speech orsinging sources. The model is optimized in an end-to-end fashion without anymanual supervision, such as lyrics, musical notes or parallel samples. Theproposed approach is fully-convolutional and can generate audio in real-time.Experiments show that our method significantly outperforms the baseline methodswhile generating convincingly better audio samples than alternative attempts.

Please select stars to rate!

         

0 comments Sign in to leave a comment.

    Data loading, please wait...
×