TTS robot conversion samples

Description

We present the conversions of a text spoken by a TTS robot to various speakers. In each row we present the conversion to the target speaker and the same text as recorded by the target speaker. The TTS speaker and the text used were not part of the method training data.

Conversions

Speaker ID TTS Generated Sample Ground Truth
4680
4088
4195
7078
4813
3259
2952
5867
669
8324
4859
78
4214
8465
887
1502
196
8838
60
6563
1088
1963
5789
3242
118
1040
4018
7800
6818
6880
5393
8580
1116
8088
8051
3486
1246
374
4788
5322
6476
2893
6836
839
6385
7190
2196
2989
6367
405
4640
587
6081
6209
7067
6064
3168
5022
2002
7402
4137
5561