audio_check
/
automatic-speech-recognition
/hw_asr
/augmentations
/spectrogram_augmentations
/TimeMasking.py
from torch import Tensor | |
from hw_asr.augmentations.base import AugmentationBase | |
from torchaudio import transforms | |
class TimeMasking(AugmentationBase): | |
def __init__(self, *args, **kwargs): | |
self._aug = transforms.TimeMasking(*args, **kwargs) | |
def __call__(self, spectogram: Tensor): | |
return self._aug(spectogram).squeeze(1) | |