Audio Item
Read more about our audio studio here.
For audio files (e.g. FLAC, mp3, etc.) we use the subtitle annotation type:
Classification
You can classify an entire audio item or by selecting start and end time. Setting the text to an empty string will not add the transcription, but only the desired class:
import dtlpy as dl
project = dl.projects.get('My Project')
dataset = project.datasets.get('My Dataset')
item = dataset.items.get(filepath='/my_item.mp3')
# Create a builder instance and load annotation from the VTT file
builder = item.annotations.builder()
builder.add(annotation_definition=dl.Subtitle(label='Whisper', text=''),
start_time=1.0,
end_time=2.5)
# Upload annotations to the item
item.annotations.upload(builder)Transcription
Our SDK supports importing VTT files directly with the from_vtt_file method:
item = dataset.items.get(filepath='/my_item.mp3')
# Create a builder instance and load annotation from the VTT file
builder = item.annotations.builder()
builder.from_vtt_file(filepath=r"E:\TypesExamples\audio\subtitles.vtt")
# Upload annotations to the item
item.annotations.upload(builder)Since there is no label in the vtt files (only time and text), the default label for those annotation will be Text.
For any other specific use, you can read the vtt file (or any other format) and upload the annotations directly:
import webvtt
import datetime
filepath = '/home/subtitles.vtt'
builder = item.annotations.builder()
for caption in webvtt.read(filepath):
h, m, s = caption.start.split(':')
start_time = datetime.timedelta(hours=float(h), minutes=float(m), seconds=float(s)).total_seconds()
h, m, s = caption.end.split(':')
end_time = datetime.timedelta(hours=float(h), minutes=float(m), seconds=float(s)).total_seconds()
builder.add(annotation_definition=dl.Subtitle(label='Speaker1',
text=caption.text),
start_time=start_time,
end_time=end_time)
item.annotations.upload(builder)