Skip to content
New issue

Have a question about this project? # for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “#”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? # to your account

wip: add VoxCeleb datasets #1

Draft
wants to merge 4 commits into
base: master
Choose a base branch
from
Draft
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
1 change: 1 addition & 0 deletions README.md
Original file line number Diff line number Diff line change
Expand Up @@ -5,6 +5,7 @@ This repository aims to centralize scripts that prepare datasets to be used with
Currently available :
- [AISHELL4](aishell4)
- [MSDWild](msdwild)
- [VoxCeleb](voxceleb)

To setup each dataset, refer to the `README.md` contained in their respective folder.

Expand Down
28 changes: 28 additions & 0 deletions voxceleb/README.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,28 @@
# VoxCeleb for pyannote.audio

## Usage

```python
from pyannote.database import registry
registry.load_database("database.yml")

protocol = registry.get_protocol("VoxCeleb.SpeakerVerification.VoxCeleb")

# iterate on all VoxCeleb1 dev (but speakers whose names starts with U, V, or W) and on all VoxCeleb2 dev
for file in protocol.train():
pass

# iterate on all VoxCeleb1 dev speakers whoe names starts with U, V, or W
for file in protocol.development():
pass

# iterate over target/non-target trials between the above UVW speakers
for trial in protocol.development_trial():
pass

# iterate over target/non-target trials defined VoxCeleb1 "original"
for trial in protocol.test_trial():
pass
```


102 changes: 102 additions & 0 deletions voxceleb/database.yml
Original file line number Diff line number Diff line change
@@ -0,0 +1,102 @@
Databases:
VoxCeleb:
- VoxCeleb1/dev/wav/{uri}.wav
- VoxCeleb1/test/wav/{uri}.wav
- VoxCeleb2/dev/aac/{uri}.wav
- VoxCeleb2/test/aac/{uri}.wav

Protocols:
VoxCeleb:
SpeakerDiarization:
# original VoxCeleb1 split (no train)
VoxCeleb1:
development:
uri: vox1.dev.lst
annotation: vox1.dev.rttm
annotated: vox1.dev.uem
duration: vox1.dev.duration.map
test:
uri: vox1.tst.lst
annotation: vox1.tst.rttm
annotated: vox1.tst.uem
duration: vox1.tst.duration.map
# original VoxCeleb2 split (no train)
VoxCeleb2:
development:
uri: vox2.dev.lst
annotation: vox2.dev.rttm
annotated: vox2.dev.uem
duration: vox2.dev.duration.map
test:
uri: vox2.tst.lst
annotation: vox2.tst.rttm
annotated: vox2.tst.uem
duration: vox2.tst.duration.map

SpeakerVerification:
# trn: all VoxCeleb1 development speakers but those whose name starts by U, V, or W.
# dev: VoxCeleb1 development speakers whose name starts by U, V, or W.
# tst: original VoxCeleb1 test set
VoxCeleb1:
train:
uri: vox1.dev_a.lst
annotation: vox1.dev.rttm
annotated: vox1.dev.uem
duration: vox1.dev.duration.map
development:
uri: vox1.dev_b.lst
annotation: vox1.dev.rttm
annotated: vox1.dev.uem
duration: vox1.dev.duration.map
trial: vox1.dev_b.trial.txt
test:
uri: vox1.tst.lst
annotation: vox1.tst.rttm
annotated: vox1.tst.uem
duration: vox1.tst.duration.map
trial: vox1.tst.trial.txt

VoxCeleb:
# trn: all VoxCeleb2 development speakers + all VoxCeleb1 speakers but those whose name starts by U, V, or W
# dev: VoxCeleb1 development speakers whose name starts by U, V, or W.
# tst: original VoxCeleb1 test set
train:
uri: vox.trn.lst
annotation: vox.trn.rttm
annotated: vox.trn.uem
duration: vox.trn.duration.map
development:
uri: vox1.dev_b.lst
annotation: vox1.dev.rttm
annotated: vox1.dev.uem
duration: vox1.dev.duration.map
trial: vox1.dev_b.trial.txt
test:
uri: vox1.tst.lst
annotation: vox1.tst.rttm
annotated: vox1.tst.uem
duration: vox1.tst.duration.map
trial: vox1.tst.trial.txt

VoxCelebMini:
# trn: First 100 speaker of VoxCeleb protocol
# dev: same as VoxCeleb.SpeakerVerification.VoxCeleb protocol
# tst: same as VoxCeleb.SpeakerVerification.VoxCeleb protocol
train:
uri: mini_vox/mini_vox.trn.lst
annotation: mini_vox/mini_vox.trn.rttm
annotated: mini_vox/mini_vox.trn.uem
duration: mini_vox/mini_vox.trn.duration.map
development:
uri: vox1.dev_b.lst
annotation: vox1.dev.rttm
annotated: vox1.dev.uem
duration: vox1.dev.duration.map
trial: vox1.dev_b.trial.txt
test:
uri: vox1.tst.lst
annotation: vox1.tst.rttm
annotated: vox1.tst.uem
duration: vox1.tst.duration.map
trial: vox1.tst.trial.txt

Loading