SpeechGenerator.py

"""
A generator for reading and serving audio files

https://stanford.edu/~shervine/blog/keras-how-to-generate-data-on-the-fly.html

Remember to use multiprocessing:
# Train model on dataset
model.fit_generator(generator=training_generator,
                    validation_data=validation_generator,
                    use_multiprocessing=True,
                    workers=6)

"""

import numpy as np
import tensorflow.keras


class SpeechGen(tensorflow.keras.utils.Sequence):
    """
    'Generates data for Keras'

    list_IDs - list of files that this generator should load
    labels - dictionary of corresponding (integer) category
    to each file in list_IDs

    Expects list_IDs and labels to be of the same length
    """
    def __init__(self, list_IDs, labels, batch_size=32,
                 dim=16000, shuffle=True):
        'Initialization'
        self.dim = dim
        self.batch_size = batch_size
        self.labels = labels
        self.list_IDs = list_IDs
        self.shuffle = shuffle
        self.on_epoch_end()

    def __len__(self):
        'Denotes the number of batches per epoch'
        return int(np.floor(len(self.list_IDs) / self.batch_size))

    def __getitem__(self, index):
        'Generate one batch of data'
        # Generate indexes of the batch
        indexes = self.indexes[index*self.batch_size:(index+1)*self.batch_size]

        # Find list of IDs
        list_IDs_temp = [self.list_IDs[k] for k in indexes]

        # Generate data
        X, y = self.__data_generation(list_IDs_temp)

        return X, y

    def on_epoch_end(self):
        'Updates indexes after each epoch'
        self.indexes = np.arange(len(self.list_IDs))
        if self.shuffle:
            np.random.shuffle(self.indexes)

    def __data_generation(self, list_IDs_temp):
        'Generates data containing batch_size samples'
        # X : (n_samples, *dim, n_channels)
        # Initialization
        X = np.empty((self.batch_size, self.dim))
        y = np.empty((self.batch_size), dtype=int)

        # Generate data
        for i, ID in enumerate(list_IDs_temp):

            # load data from file, saved as numpy array on disk
            curX = np.load(ID)[:, 0]

            # normalize
            # invMax = 1/(np.max(np.abs(curX))+1e-3)
            # curX *= invMax

            # curX could be bigger or smaller than self.dim
            if curX.shape[0] == self.dim:
                X[i] = curX
            elif curX.shape[0] > self.dim:  # bigger
                # we can choose any position in curX-self.dim
                randPos = np.random.randint(curX.shape[0]-self.dim)
                X[i] = curX[randPos:randPos+self.dim]
            else:  # smaller
                randPos = np.random.randint(self.dim-curX.shape[0])
                X[i, randPos:randPos + curX.shape[0]] = curX
                # print('File dim smaller')

            # Store class
            y[i] = self.labels[ID]

        return X, y