This website contains supplementary material
for our paper
``A bottleneck auto-encoder for F0 transformations on speech and singing voice'',
currently under review for the
Special Issue - Signal Processing with Convolutional Neural Networks
of the MDPI journal Information.
Samples for Singing Voice
Below are samples from the test set of our singing dataset.
Click on a sample to open a summary page for that sample,
containing transpositions from different models.
Soprano
1-2-2
,
1-2-5
,
1-2-19
,
1-2-154
,
1-2-155
,
Alto
9-4-62
,
9-4-64
,
9-4-67
,
9-4-73
,
9-4-78
,
Counter Tenor
2-4-51
,
2-4-77
,
2-4-92
,
2-4-158
,
Tenor
2-3-45
,
2-3-87
,
2-3-122
,
2-3-140
,
2-3-142
,
Baritone
1-1-332
,
1-1-337
,
1-1-366
,
1-1-390
,
1-1-404
,
Byzantine
3-41
,
3-76
,
3-120
,
3-172
,
3-233
,
Childvoice
2-1-86
,
2-1-136
,
2-1-168
,
2-1-186
,
2-1-192
,
French Pop Female
4-1-1405
,
4-1-1480
,
4-1-1500
,
4-1-1508
,
4-1-1514
,
JPop Male
7-0-21
,
7-0-35
,
7-0-48
,
7-0-71
,
7-0-97
,
JPop Female
8-1-0-22
,
8-1-0-56
,
8-1-0-75
,
8-1-0-96
,
8-1-0-113
,
unseen female
9-2-74
,
9-2-76
,
9-2-77
,
9-2-78
,
9-2-79
,
9-2-80
,
9-2-81
,
9-2-84
,
9-2-86
,
unseen male
9-15-80
,
9-15-81
,
9-15-82
,
9-15-83
,
9-15-84
,
9-15-85
,
9-15-86
,
9-15-87
,
9-15-88
,
Samples for Speech
Below are samples from the test set of our speech dataset.
Click on a sample to open a summary page for that sample,
containing transpositions from different models.
VCTK Female p361
10-205-1
,
10-205-2
,
10-205-3
,
10-205-4
,
10-205-5
,
10-205-6
,
10-205-7
,
10-205-8
,
VCTK Female p362
10-207-1
,
10-207-2
,
10-207-3
,
10-207-4
,
10-207-5
,
10-207-6
,
10-207-7
,
10-207-8
,
VCTK Male p374
10-213-1
,
10-213-2
,
10-213-3
,
10-213-4
,
10-213-5
,
10-213-6
,
10-213-7
,
10-213-8
,
VCTK Male p376
10-215-1
,
10-215-2
,
10-215-3
,
10-215-4
,
10-215-5
,
10-215-6
,
10-215-7
,
10-215-8
,