mirror of
https://github.com/pyannote/pyannote-audio.git
synced 2026-11-20 13:07:51 +09:00
8.3 KiB
8.3 KiB
In [ ]:
# TODO: switch to AMI
PROTOCOL = 'Debug.SpeakerDiarization.Debug'In [ ]:
from pyannote.database import get_protocol, FileFinder
protocol = get_protocol(PROTOCOL, preprocessors={"audio": FileFinder()})In [ ]:
first_training_file = next(protocol.train())In [ ]:
from pyannote.audio.utils.preview import listen
listen(first_training_file)In [ ]:
first_training_file['annotation']In [ ]:
from pyannote.audio.pipelines.voice_activity_detection import OracleVoiceActivityDetection
oracle_vad = OracleVoiceActivityDetection()
oracle_vad(first_training_file).get_timeline()In [ ]:
from pyannote.audio.tasks import VoiceActivityDetection
vad = VoiceActivityDetection(protocol, duration=2., batch_size=16)In [ ]:
from pyannote.audio.models.debug import SimpleSegmentationModel
model = SimpleSegmentationModel(task=vad)In [ ]:
import pytorch_lightning as pl
trainer = pl.Trainer(max_epochs=10)
trainer.fit(model, vad)In [ ]:
test_file = next(protocol.test())
# here we use a test file provided by the protocol, but it could be any audio file
# e.g. test_file = "/path/to/test.wav".In [ ]:
from pyannote.audio.core.inference import Inference
inference = Inference(model)
vad_probability = inference(test_file)In [ ]:
vad_probabilityIn [ ]:
from pyannote.audio.pipelines import VoiceActivityDetection as VoiceActivityDetectionPipeline
pipeline = VoiceActivityDetectionPipeline(scores=inference).instantiate(
{"onset": 0.5, "offset": 0.5, "min_duration_on": 0.1, "min_duration_off": 0.1})In [ ]:
pipeline(test_file).get_timeline()In [ ]:
# to make things faster, we run the inference once and for all...
validation_files = list(protocol.development())
for file in validation_files:
file['vad_scores'] = inference(file)
# ... and tell the pipeline to load VAD scores directly from files
pipeline = VoiceActivityDetectionPipeline(scores="vad_scores")In [ ]:
from pyannote.pipeline import Optimizer
optimizer = Optimizer(pipeline)
optimizer.tune(validation_files, n_iterations=200, show_progress=False)In [ ]:
optimized_pipeline = VoiceActivityDetectionPipeline(scores=inference).instantiate(optimizer.best_params)
optimized_pipeline(test_file).get_timeline()