Tacotron: A Fully End-to-End Text-To-Speech Synthesis Model

Re:Tacotron: A Fully End-to-End Text-To-Speech Synthesis Model by yunolisten on Thursday March 30, 2017 @05:22AM (#13993962)

I'm totally new to tts, so please forgive the questions:

* Is this open source?

* What hardware would one need to run this?

* If specialist hardware is required (e.g. reasonably high end GPU) would that simply be for building and training the system or its operation?

[ Parent | on HN ]

Re:Tacotron: A Fully End-to-End Text-To-Speech Synthesis Model by asadjb on Thursday March 30, 2017 @07:01AM (#13994361)

From what I understand looking through the paper, this is only a research paper, without any source code. So right now it's not possible to say what the hardware requirement might be.

[ Parent | on HN ]

Re:Tacotron: A Fully End-to-End Text-To-Speech Synthesis Model by yunolisten on Thursday March 30, 2017 @07:27AM (#13994455)

Thank you. It would have made more sense to me if the source code to generate the output examples which they posted. Oh well.

[ Parent | on HN ]