athena.transform¶
Subpackages¶
athena.transform.featsathena.transform.feats.opsathena.transform.feats.base_frontendathena.transform.feats.cmvnathena.transform.feats.cmvn_testathena.transform.feats.fbankathena.transform.feats.fbank_pitchathena.transform.feats.fbank_pitch_testathena.transform.feats.fbank_testathena.transform.feats.framepowathena.transform.feats.framepow_testathena.transform.feats.mel_spectrumathena.transform.feats.mel_spectrum_testathena.transform.feats.mfccathena.transform.feats.mfcc_testathena.transform.feats.pitchathena.transform.feats.pitch_testathena.transform.feats.read_wavathena.transform.feats.read_wav_testathena.transform.feats.spectrumathena.transform.feats.spectrum_testathena.transform.feats.write_wavathena.transform.feats.write_wav_test
Submodules¶
Package Contents¶
Classes¶
AudioFeaturizer |
Interface of audio Features extractions. |
Functions¶
compute_cmvn(audio_feature, mean=None, variance=None, local_cmvn=False) |
Compute cmvn on feature. |
read_wav(wavfile, audio_channels=1) |
read wav from file |
-
class
athena.transform.AudioFeaturizer(config={'type': 'Fbank'})¶ Interface of audio Features extractions.
-
dim¶ return the dimension of the feature if only ReadWav, return 1
-
num_channels¶ return the channel of the feature
-
__call__(self, audio=None, sr=None, speed=1.0)¶ extract feature from audo data :param audio data or audio file :sr sample rate :return feature
-
__impl(self, audio=None, sr=None, speed=1.0)¶ :param audio data or audio file, a tensor :sr sample rate, a tensor :return feature
-
-
athena.transform.compute_cmvn(audio_feature, mean=None, variance=None, local_cmvn=False)¶ Compute cmvn on feature.
-
athena.transform.read_wav(wavfile, audio_channels=1)¶ read wav from file args: audio_channels = 1 returns: tf.squeeze(audio_data * 32768, axis=-1), tf.cast(sample_rate, dtype=tf.int32)