paddlespeech.s2t.frontend.augmentor.impulse_response module

Contains the impulse response augmentation model.

class paddlespeech.s2t.frontend.augmentor.impulse_response.ImpulseResponseAugmentor(rng, impulse_manifest_path)[source]

Bases: AugmentorBase

Augmentation model for adding impulse response effect.

Parameters:

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

  • impulse_manifest_path (str) -- Manifest path for impulse audio data.

Methods

__call__(x[, uttid, train])

Call self as a function.

transform_audio(audio_segment)

Add impulse response effect.

transform_feature(spec_segment)

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

Add impulse response effect.

Note that this is an in-place transformation.

Parameters:

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