SegmentSpan.md

SegmentSpan

One ASR segment's span inside a media chunk (media chunks only). Field names are deliberately compact — a media document stores hundreds of chunks x tens of spans each in chunk_metadata JSONB. Measured worst case is a 4-hour Mandarin recording: 175 chunks x ~33 spans, ~190 KB total.

Properties

Name	Type	Description	Notes
s	int	Segment start in the media, ms from start.
e	int	Segment end in the media, ms from start.
c	int	Character offset of this segment's text within the chunk's joined content — maps a position in the chunk text back to a moment in the recording.
k	int	Speaker label for this segment, 0-indexed within the recording. It identifies turns, not people — there is no voiceprint enrollment, so label 0 in one meeting is unrelated to label 0 in another. None when the ASR provider does not diarize.	[optional]

Example

from ksapi.models.segment_span import SegmentSpan

# TODO update the JSON string below
json = "{}"
# create an instance of SegmentSpan from a JSON string
segment_span_instance = SegmentSpan.from_json(json)
# print the JSON string representation of the object
print(SegmentSpan.to_json())

# convert the object into a dict
segment_span_dict = segment_span_instance.to_dict()
# create an instance of SegmentSpan from a dict
segment_span_from_dict = SegmentSpan.from_dict(segment_span_dict)

[Back to Model list] [Back to API list] [Back to README]