|
@ -23,8 +23,6 @@ import resampy |
|
|
import mel_features |
|
|
import mel_features |
|
|
import vggish_params |
|
|
import vggish_params |
|
|
|
|
|
|
|
|
import torchaudio |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
def waveform_to_examples(data, sample_rate, return_tensor=True): |
|
|
def waveform_to_examples(data, sample_rate, return_tensor=True): |
|
|
"""Converts audio waveform into an array of examples for VGGish. |
|
|
"""Converts audio waveform into an array of examples for VGGish. |
|
@ -78,22 +76,3 @@ def waveform_to_examples(data, sample_rate, return_tensor=True): |
|
|
log_mel_examples, requires_grad=True)[:, None, :, :].float() |
|
|
log_mel_examples, requires_grad=True)[:, None, :, :].float() |
|
|
|
|
|
|
|
|
return log_mel_examples |
|
|
return log_mel_examples |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
def wavfile_to_examples(wav_file, return_tensor=True): |
|
|
|
|
|
""" |
|
|
|
|
|
Convenience wrapper around waveform_to_examples() for a common WAV format. |
|
|
|
|
|
|
|
|
|
|
|
Args: |
|
|
|
|
|
wav_file: |
|
|
|
|
|
String path to a file, or a file-like object. |
|
|
|
|
|
The file is assumed to contain WAV audio data with signed 16-bit PCM samples. |
|
|
|
|
|
return_tensor: |
|
|
|
|
|
Return data as a Pytorch tensor ready for VGGish |
|
|
|
|
|
|
|
|
|
|
|
Returns: |
|
|
|
|
|
See waveform_to_examples. |
|
|
|
|
|
""" |
|
|
|
|
|
data, sr = torchaudio.load(wav_file) |
|
|
|
|
|
wav_data = data.detach().numpy().transpose() |
|
|
|
|
|
return waveform_to_examples(wav_data, sr, return_tensor) |
|
|
|
|
|