Expressive Neural Voice Cloning Demo

Please record audio for the following texts by pressing the Record and Stop buttons. Try to be as accurate as possible while reading the texts and avoid silences in the beginning and at the end of a recording. You may re-record a particular text any number of times. When you are done with recording all the texts, press the upload button. After the upload is complete, you will be able to hear the results of zeroshot voice cloning and model adaptation will start in the background.

SNo. Transcript Record Stop Recorded Audio
1Police were called to the scene, according to the weekend reports.
2There was a tremendous amount of anger and frustration in that music.
3It would still have been a good film, but very different.
4That review will look at management across the sector as a whole.
5Naturally, it was not difficult to find support for these proposals.
6Since backboned animals are best known to most readers, they may be taken as an illustration.
7The rainbow is a division of white light into many beautiful colors.
8Out beyond ideas of wrongdoing and rightdoing there is a field, I'll meet you there.
9He also has property interests in the country.
10The committee will meet this afternoon for a special debate.
Log


			
Recordings
    Model Adaptation Progress:
    0%
    Style Transfer
    Reference Type Style Reference Audio Synthesized - Zeroshot Synthesized - Finetuned
    Input Text Zeroshot Finetuned

      Launch a model server using our google colab notebook (should take a couple of minutes to setup) and paste the address here.