Zero-shot test on Indian-English accent to general American-English accent

        Samples from VCTK

Transcript Source accent Liu. et al Generative model(EnCodec) Generative model(TF-Codec) The proposed
She is given a new deputy minister for transport and planning.
I had relied on him.
There is a lack of chemistry.
It wasn't to be.
It's just awful.
It is set in Paris.
It is also very valuable.
We must provide a long-term solution to tackle this attitude.
To do so he reckons that a good opening result is essential.
We also need a small plastic snake and a big toy frog for the kids.

        Samples from L1-L2 ARCTIC

Transcript Source accent Generative model(EnCodec) Generative model(TF-Codec) The proposed
There was nothing for us but the wide raw ocean.
There were orange green gold green and a copper green.
The boy hesitated then mastered his temper.
Too much he told me with ominous rolling head.
He had become a man very early in life.
Also she wouldn't walk.
He was manifestly distressed by my coming.
You're going in for grab sharing.
Eighteen he added.
He was an amphibian and a mountaineer.

Zero-shot tests on more source accents

        Source accent: Chinese-English -> Target accent: General American-English

Transcript Source accent Generative model(EnCodec) Generative model(TF-Codec) The proposed
Author of the danger trail Philip Steels etc.
Now you're coming down to business Phil he exclaimed.
I'm playing a single hand in what looks like a losing game.
I suppose you wonder why she is coming up here.

        Source accent: Korean-English -> Target accent: General American-English

Transcript Source accent Generative model(EnCodec) Generative model(TF-Codec) The proposed
I was about to do this when cooler judgment prevailed.
And now down there Eileen was waiting for him.
Their forces were already moving into the north country.
I suppose you wonder why she is coming up here.

A/B testing samples in Section 4.5.2

        Source of semantic tokens: utterance in general American-English

Transcript Content source American-English prompt Generative model(TF-Codec) Indian-English prompt Generative model(TF-Codec)
The rainbow is a division of white light into many beautiful colors.
There is , according to legend, a boiling pot of gold at one end.
People look, but no one ever finds it.
The Greeks used to imagine that it was a sign from the gods to foretell war or heavy rain.