paddlespeech.s2t.frontend.augmentor.online_bayesian_normalization module

Contain the online bayesian normalization augmentation model.

class paddlespeech.s2t.frontend.augmentor.online_bayesian_normalization.OnlineBayesianNormalizationAugmentor(rng, target_db, prior_db, prior_samples, startup_delay=0.0)[source]

Bases: AugmentorBase

Augmentation model for adding online bayesian normalization.

Parameters:

  • rng (random.Random) -- Random generator object.

  • target_db (float) -- Target RMS value in decibels.

  • prior_db (float) -- Prior RMS estimate in decibels.

  • prior_samples (int) -- Prior strength in number of samples.

  • startup_delay -- Default 0.0s. If provided, this function will accrue statistics for the first startup_delay seconds before applying online normalization.

Methods

__call__(x[, uttid, train])

Call self as a function.

transform_audio(audio_segment)

Normalizes the input audio using the online Bayesian approach.

transform_feature(spec_segment)

Adds various effects to the input audo feature segment.
transform_audio(audio_segment)[source]

Normalizes the input audio using the online Bayesian approach.

Note that this is an in-place transformation.

Parameters:

audio_segment (AudioSegment|SpeechSegment) -- Audio segment to add effects to.