Codes and notes for learning audio operations with pytorch

Repo Info

• Related Series in YouTube: PyTorch for Audio + Music Processing

  Codes given by the author: GitHub Repo

• Files / Directories Info (Listed in the order explained in the tutorial above):

  Directories:

  • MNIST: A dataset, auto downloaded in file train.py.
  • UrbanSound8K: A dataset, downloaded from website URBANSOUND8K DATASET

  Files:

  • train_feed_forward_network.py: Contains a class FeedForwardNet and functions download_mnist_datasets, train_one_epoch and train which are used for downloading MNIST dataset and training them using FeedForwardNet model.
  • feedforwardnet.pth: Model saved from train.py.
  • predict_feed_forward_network.py: Contains a function predict for validating the model feedforwardnet.pth.
  • urban_sound_dataset.py: Contains a class UrbanSoundDataset for loading .wav sound file in urbansound8k dataset and getting the waveform signals, sample rates and mel-spectorgrams of each audio. Serveral works are done in method __getitem__:
    • Load the .wav audio file and get its waveform signal and sample rate.
    • Resample the signal if the original sample rate is not equal to the target sample rate.
    • Mix down multiple channels to moto.
    • If the number of samples is more than the expected, apply cutting operation.
    • if the number of samples is less than the expected, apply right padding operation.
    • Use transforming function (here it's mel_spectrogram) to transform it.
  • cnn.py: A simple CNN model.
  • train_cnn: Use the model in cnn.py to train the urban sound dataset.
  • predict_cnn: Predict the model in the same way of predict_feed_forward_network.py.

Some Environmental Problems

• Get RuntimeError: No audio I/O backend is available. message while running code torchaudio.load(audio_sample_path) at file urban_sound_dataset.py:

  # try with commands below
  pip install SoundFile
  # or
  pip install sox

• Get error message below when plotting mel-spectrogram using matplotlib:

   manager_pyplot_show = vars(manager_class).get("pyplot_show")
  TypeError: vars() argument must have __dict__ attribute
  

  Solutions (Stack Overflow):

  mpl.use('TkAgg')  # Add this code