.. DO NOT EDIT.
.. THIS FILE WAS AUTOMATICALLY GENERATED BY SPHINX-GALLERY.
.. TO MAKE CHANGES, EDIT THE SOURCE PYTHON FILE:
.. "acoustic/acoustic_spectrogram_aug_tutorial.py"
.. LINE NUMBERS ARE GIVEN BELOW.

.. only:: html

    .. note::
        :class: sphx-glr-download-link-note

        :ref:`Go to the end <sphx_glr_download_acoustic_acoustic_spectrogram_aug_tutorial.py>`
        to download the full example code.

.. rst-class:: sphx-glr-example-title

.. _sphx_glr_acoustic_acoustic_spectrogram_aug_tutorial.py:


Acoustic augmentation.spectrogram_aug Tutorial
==============================================================

.. GENERATED FROM PYTHON SOURCE LINES 7-8

!pip install pysensing

.. GENERATED FROM PYTHON SOURCE LINES 10-12

In this tutorial, we will be implementing a simple acoustic.augmentation.spectrogram_aug


.. GENERATED FROM PYTHON SOURCE LINES 12-21

.. code-block:: Python

    import torch
    import librosa
    import numpy as np
    import torchaudio
    import matplotlib.pyplot as plt
    import sys
    import pysensing.acoustic.augmentation.spectrogram_aug as spec_aug
    import pysensing.acoustic.preprocessing.transform as transform








.. GENERATED FROM PYTHON SOURCE LINES 22-25

Load the audio
-----------------------------------
First, the example audio is loaded

.. GENERATED FROM PYTHON SOURCE LINES 25-47

.. code-block:: Python


    # Define the plot function
    def plot(specs, titles):
        maxlen = max(spec.shape[-1] for spec in specs)
    
        def plot_spec(ax, spec, title):
            ax.set_title(title)
            ax.imshow(spec, origin="lower", aspect="auto")
            ax.set_xlim(0, maxlen)
    
        num_specs = len(specs)
        fig, axes = plt.subplots(num_specs, 1,figsize=(12,8))
        if num_specs == 1:
            axes = [axes]
        for ax, spec, title in zip(axes, specs, titles):
            plot_spec(ax, spec[0].float(), title)
        fig.tight_layout()
        plt.show()

    # Load the data
    waveform, sample_rate = torchaudio.load('example_data/example_audio.wav')
    spectrogram = transform.spectrogram()(waveform)







.. GENERATED FROM PYTHON SOURCE LINES 48-50

1. Timestretch
------------------------

.. GENERATED FROM PYTHON SOURCE LINES 50-60

.. code-block:: Python


    # Define timestretch with different fixed_rate
    timestretch_compress = spec_aug.timestretch(fixed_rate=1.1)
    timestretch_extend   = spec_aug.timestretch(fixed_rate=0.9)
    # Do timestretch to the input spectrogram
    spectrogram_com = timestretch_compress(spectrogram)
    spectrogram_ext = timestretch_extend(spectrogram)
    # Plotting
    plot([spectrogram,spectrogram_com,spectrogram_ext],['Original','Fixed_rate=1.1','Fixed_rate=1.9'])




.. image-sg:: /acoustic/images/sphx_glr_acoustic_spectrogram_aug_tutorial_001.png
   :alt: Original, Fixed_rate=1.1, Fixed_rate=1.9
   :srcset: /acoustic/images/sphx_glr_acoustic_spectrogram_aug_tutorial_001.png
   :class: sphx-glr-single-img


.. rst-class:: sphx-glr-script-out

 .. code-block:: none

    /home/kemove/yyz/av-gihub/tutorials/acoustic_source/acoustic_spectrogram_aug_tutorial.py:40: UserWarning: Casting complex values to real discards the imaginary part (Triggered internally at ../aten/src/ATen/native/Copy.cpp:299.)
      plot_spec(ax, spec[0].float(), title)




.. GENERATED FROM PYTHON SOURCE LINES 61-63

2. Timemasking
------------------------

.. GENERATED FROM PYTHON SOURCE LINES 63-71

.. code-block:: Python

    timemasking_trans          = spec_aug.timemasking(200)
    timemasking_random_trans   = spec_aug.timemasking(200,p=0.5)

    timemask_spec   = timemasking_trans(spectrogram)
    timemask_r_spec = timemasking_random_trans(spectrogram)

    plot([spectrogram,timemask_spec,timemask_r_spec],['Original','Timemasking','Timemasking_random'])




.. image-sg:: /acoustic/images/sphx_glr_acoustic_spectrogram_aug_tutorial_002.png
   :alt: Original, Timemasking, Timemasking_random
   :srcset: /acoustic/images/sphx_glr_acoustic_spectrogram_aug_tutorial_002.png
   :class: sphx-glr-single-img





.. GENERATED FROM PYTHON SOURCE LINES 72-74

3. Frequecymasking
------------------------

.. GENERATED FROM PYTHON SOURCE LINES 74-79

.. code-block:: Python


    frequencymasking_trans          = spec_aug.frequencymasking(400)
    frequencymask_spec   = frequencymasking_trans(spectrogram)
    plot([spectrogram,frequencymask_spec],['Original','Frequencymasking'])




.. image-sg:: /acoustic/images/sphx_glr_acoustic_spectrogram_aug_tutorial_003.png
   :alt: Original, Frequencymasking
   :srcset: /acoustic/images/sphx_glr_acoustic_spectrogram_aug_tutorial_003.png
   :class: sphx-glr-single-img





.. GENERATED FROM PYTHON SOURCE LINES 80-81

And that's it. We're done with our acoustic augmentation.signal_aug tutorials. Thanks for reading.


.. rst-class:: sphx-glr-timing

   **Total running time of the script:** (0 minutes 0.429 seconds)


.. _sphx_glr_download_acoustic_acoustic_spectrogram_aug_tutorial.py:

.. only:: html

  .. container:: sphx-glr-footer sphx-glr-footer-example

    .. container:: sphx-glr-download sphx-glr-download-jupyter

      :download:`Download Jupyter notebook: acoustic_spectrogram_aug_tutorial.ipynb <acoustic_spectrogram_aug_tutorial.ipynb>`

    .. container:: sphx-glr-download sphx-glr-download-python

      :download:`Download Python source code: acoustic_spectrogram_aug_tutorial.py <acoustic_spectrogram_aug_tutorial.py>`


.. only:: html

 .. rst-class:: sphx-glr-signature

    `Gallery generated by Sphinx-Gallery <https://sphinx-gallery.github.io>`_