如何将mfcc矢量与注释中的标签结合起来传递给神经网络

时间:2018-01-22 19:07:34

标签: python neural-network keras mfcc librosa

使用librosa,我为我的音频文件创建了mfcc,如下所示:

import librosa
y, sr = librosa.load('myfile.wav')
print y
print sr
mfcc=librosa.feature.mfcc(y=y, sr=sr)

我还有一个文本文件,其中包含与音频对应的手动注释[start,stop,tag],如下所示:

  

0.0 2.0 sound1
  2.0 4.0 sound2
  4.0 6.0沉默
  6.0 8.0 sound1

问题: 如何将生成的librosa生成的mfcc与文本文件中的注释结合起来。

最终目标是,我想结合对应于标签的mfcc,并传递
 它到神经网络。
因此,神经网络将mfcc和相应的标签作为训练数据。

如果它是一维的,我可以有N列N值,最后一列Y带有Class标签。 但我很困惑如何继续,因为mfcc具有类似的形状 (16,X)或 (20,Y)。 所以我不知道如何将两者结合起来。

我的示例mfcc位于:https://gist.github.com/manbharae/0a53f8dfef6055feef1d8912044e1418

请帮助谢谢。

更新:目标是训练神经网络,以便在将来遇到它时识别出新的声音。

我用Google搜索,发现mfcc非常适合演讲。然而,我的音频有语音,但我想识别非语音。是否有其他推荐的音频功能用于通用音频分类/识别任务?

1 个答案:

答案 0 :(得分:5)

尝试以下方法。解释包含在代码中。

import numpy
import librosa

# The following function returns a label index for a point in time (tp)
# this is psuedo code for you to complete
def getLabelIndexForTime(tp):
    # search the loaded annoations for what label corresponsons to the given time
    # convert the label to an index that represents its unqiue value in the set
    # ie.. 'sound1' = 0, 'sound2' = 1, ...
    #print tp  #for debug
    label_index = 0 #replace with logic above
    return label_index


if __name__ == '__main__':
    # Load the waveforms samples and convert to mfcc
    raw_samples, sample_rate = librosa.load('Front_Right.wav')
    mfcc  = librosa.feature.mfcc(y=raw_samples, sr=sample_rate)
    print 'Wave duration is %4.2f seconds' % (len(raw_samples)/float(sample_rate))

    # Create the network's input training data, X
    # mfcc is organized (feature, sample) but the net needs (sample, feature)
    # X is mfcc reorganized to (sample, feature)
    X     = numpy.moveaxis(mfcc, 1, 0)
    print 'mfcc.shape:', mfcc.shape
    print 'X.shape:   ', X.shape

    # Note that 512 samples is the default 'hop_length' used in calculating 
    # the mfcc so each mfcc spans 512/sample_rate seconds.
    mfcc_samples = mfcc.shape[1]
    mfcc_span    = 512/float(sample_rate)
    print 'MFCC calculated duration is %4.2f seconds' % (mfcc_span*mfcc_samples)

    # for 'n' network input samples, calculate the time point where they occur
    # and get the appropriate label index for them.
    # Use +0.5 to get the middle of the mfcc's point in time.
    Y = []
    for sample_num in xrange(mfcc_samples):
        time_point = (sample_num + 0.5) * mfcc_span
        label_index = getLabelIndexForTime(time_point)
        Y.append(label_index)
    Y = numpy.array(Y)

    # Y now contains the network's output training values
    # !Note for some nets you may need to convert this to one-hot format
    print 'Y.shape:   ', Y.shape
    assert Y.shape[0] == X.shape[0] # X and Y have the same number of samples

    # Train the net with something like...
    # model.fit(X, Y, ...   #ie.. for a Keras NN model

我应该提到这里Y数据旨在用于具有softmax输出的网络,该输出可以使用整数标签数据进行训练。 Keras模型接受sparse_categorical_crossentropy损失函数(我相信损失函数在内部将其转换为单热编码)。其他框架要求Y训练标签以一热编码格式进行传播。这种情况比较常见。有很多关于如何进行转换的例子。对于你的情况,你可以做一些像......

Yoh = numpy.zeros(shape=(Y.shape[0], num_label_types), dtype='float32')
for i, val in enumerate(Y):
    Yoh[i, val] = 1.0

至于mfcc对于非语音分类是可以接受的,我希望它们可以工作,但你可能想尝试修改它们的参数,即.. librosa允许你做n_mfcc=40这样的事情所以你获得40个功能而不是20个。为了好玩,您可以尝试使用相同大小的简单FFT替换mfcc(512个样本)并查看哪个功能最佳。