Audio samples generated by the code in the t04glovern/gst-tacotron repo, which is a Tensorflow implementation of the Style Tokens: Unsupervised Style Modeling, Control and Transfer in End-to-End Speech Synthesis and Towards End-to-End Prosody Transfer for Expressive Speech Synthesis with Tacotron

• The first column is generated speech with global style token(GST) after 503K steps, which uses the target speech (column 2) as reference audio.
• The second column is arget reference audio (if exists).

1.The masters must look to that, we can't.

With GST (using target reference)	target reference audio

2. This is not the first time, nor the second, but it shall be the last.

With GST (using target reference)	target reference audio

3. It is a very anxious period for her.

With GST (using target reference)	target reference audio

4. At any other time this would have been felt dreadfully.

With GST (using target reference)	target reference audio

5. You have not the smallest chance of moving me.

With GST (using target reference)	target reference audio

6. A tensorflow implementation of the Style Tokens Unsupervised Style Modeling, Control and Transfer in End-to-End Speech Synthesis.

With GST (using target reference)	target reference audio