TRANSFER THE LINGUISTIC REPRESENTATIONS FROM TTS TO ACCENT CONVERSION WITH NON-PARALLEL DATA

Xi Chen, Jiakun Pei, Liumeng Xue, Mingyang Zhang

School of Data Science, The Chinese University of Hong Kong, Shenzhen(CUHK-Shenzhen), China

Notes

The L1 speaker speaks the General American Accent
ASI is a male speaker with Hindi accent
SVBI is a female speaker with Hindi accent
Dataset (L2-ARCTIC corpus [1]): https://psi.engr.tamu.edu/l2-arctic-corpus/

Experiment :

L2 speaker	Text	Input speech	TTS	Baseline	Proposed	Ablation-1	Ablation-2
ASI	I did not think you would be so early.
ASI	By virtue of that power we shall remain in power.
SVBI	I did not think you would be so early.
SVBI	He was a merry monarch, especially so for an asiatic.

References

[1] G. Zhao et al., "L2-ARCTIC: A non-native English speech corpus," in Proc. Interspeech, 2018, pp. 2783-2787.