def to_overlap(reference: Annotation) -> Annotation:
"""Get overlapped speech reference annotation
Parameters
----------
reference : Annotation
File yielded by pyannote.database protocols.
Returns
-------
overlap : `pyannote.core.Annotation`
Overlapped speech reference.
"""
overlap = Timeline(uri=reference.uri)
for (s1, t1), (s2, t2) in reference.co_iter(reference):
l1 = reference[s1, t1]
l2 = reference[s2, t2]
if l1 == l2:
continue
overlap.add(s1 & s2)
return overlap.support().to_annotation()
if not pyannote_ocr:
named_speakers = Annotation(uri=uri)
else:
named_speakers = argmax_tagger(pyannote_ocr, pyannote_speakers)
named_speakers = named_speakers.subset(pyannote_ocr.labels())
path = FUSION.format(repository=REPOSITORY, uri=uri)
directory = os.path.dirname(path)
if not os.path.exists(directory):
os.makedirs(directory)
with open(path, 'w') as fp:
duplicates = dict()
for (speech_turn, track), (_, shot_id) in named_speakers.co_iter(pyannote_shots):
original_person_name = named_speakers[speech_turn, track]
person_name = mapping.setdefault(
original_person_name, get_valid_person_name(original_person_name))
if person_name not in evidences:
segment, _ = list(itertools.islice(pyannote_ocr.subset([original_person_name]).itertracks(), 1))[0]
evidences[person_name] = {
'person_name': person_name,
'corpus_id': corpus_id,
'video_id': video_id,
'modality': 'written',
'timestamp': segment.middle
}
if not pyannote_ocr:
named_face = Annotation(uri=uri)
else:
named_face = argmax_tagger(pyannote_ocr, pyannote_face)
named_face = named_face.subset(pyannote_ocr.labels())
path = FUSION.format(repository=REPOSITORY, uri=uri)
directory = os.path.dirname(path)
if not os.path.exists(directory):
os.makedirs(directory)
with open(path, 'w') as fp:
duplicates = dict()
for (segment, track_id), (_, shot_id) in named_face.co_iter(pyannote_shots):
original_person_name = named_face[segment, track_id]
person_name = mapping.setdefault(
original_person_name, get_valid_person_name(original_person_name))
if person_name not in evidences:
segment, _ = list(itertools.islice(pyannote_ocr.subset([original_person_name]).itertracks(), 1))[0]
evidences[person_name] = {
'person_name': person_name,
'corpus_id': corpus_id,
'video_id': video_id,
'modality': 'written',
'timestamp': segment.middle
}
