nemo_curator.stages.audio.tagging.utils

Module Contents

Functions

Name	Description
`add_non_speaker_segments`	Add non-speaker segments to the segments list with speaker id ‘no-speaker’.

API

nemo_curator.stages.audio.tagging.utils.add_non_speaker_segments(
    segments: list[dict[str, typing.Any]],
    audio_duration: float,
    max_length: float | None = None
) -> None

Add non-speaker segments to the segments list with speaker id ‘no-speaker’.

If max_length is provided, splits non-speaker regions into chunks of that length; otherwise adds one segment per gap. Modifies segments in-place and sorts by start time.

Parameters:

segments

list[dict[str, Any]]

List of segment dicts with ‘start’ and ‘end’.

audio_duration

float

Total audio duration in seconds.

max_length

float | NoneDefaults to None

Optional max length for each non-speaker segment.