Speaker de-identification using diphone recognition and speech synthesis thumbnail
Pause
Mute
Subtitles not available
Playback speed
0.25
0.5
0.75
1
1.25
1.5
1.75
2
Full screen

Speaker de-identification using diphone recognition and speech synthesis

Published on Jul 02, 20151058 Views

Related categories

Chapter list

Speaker De-identification using Diphone Recognition and Speech Synthesis00:00
Presentation overview00:42
DROPSY - Diphone Recognition and Speech Synthesis System01:20
DROPSY02:39
System evaluation and results04:51
Intelligibility assessment - 105:21
Intelligibility assessment - 206:15
Intelligibility assessment - 307:03
Intelligibility assessment - 408:37
Intelligibility assessment - 508:52
Intelligibility assessment - 609:44
De-identification efficacy10:42
Baseline performance of speaker recognition system11:49
Efficiency of the de-identification procedure - 112:30
Efficiency of the de-identification procedure - 213:16
Shortcomings, improvements and further experiments16:12
Conclusions17:27
Thank you18:10