Basic Examples

Retrieve the resource:

In[1]:=

Out[2]=

Retrieve the dataset:

In[3]:=

Out[4]=

Visualization

Select an entry from the dataset:

In[5]:=

Out[6]=

Visualize the waveform of the Audio object:

In[7]:=

Out[7]=

Visualize the spectrum:

In[8]:=

Out[8]=

Visualize the spectrogram:

In[9]:=

Out[9]=

Analysis

Compute the number of examples per class:

In[10]:=

Out[11]=

Inspect the sample rate and channel count of the Audio objects:

In[12]:=

Out[13]=

Plot the histogram of the durations of the Audio objects:

In[14]:=

Out[15]=

By computing the cumulative duration of the labels, it can be seen that the dataset is not balanced:

In[16]:=

Out[17]=

The average recording duration for the two labels is different as well:

In[18]:=

Out[19]=

Create a balanced training dataset.

In[20]:=

Make sure that the class total duration of the selected samples is similar:

In[21]:=

Out[21]=

Out[22]=

Join the two classes and shuffle the elements:

In[23]:=

Out[24]=

Create an unbalanced testing dataset using the remaining data:

In[25]:=

In[26]:=

Out[26]=