python提取mfcc特征

理论部分请看：https://haythamfayek.com/2016/04/21/speech-processing-for-machine-learning.html
这是我认为最完整的、讲解最清楚的python提取mfcc特征的教程。
用到的OSR_us_000_0010_8k.wav数据在这里下载：
http://www.voiptroubleshooter.com/open_speech/american.html
完整Python代码如下：

import numpy as np
import scipy.io.wavfile
import matplotlib.pylab as plt
sample_rate, signal = scipy.io.wavfile.read("OSR_us_000_0010_8k.wav")
signal = signal[0: int(3.5*sample_rate)]
t = np.linspace(0, 3.5, num=len(signal))

pre_emphasis = 0.97
emphasized_signal = np.append(signal[0], signal[1:] - pre_emphasis*signal[:-1])
frame_size = 0.025
frame_stride = 0.01
frame_length, frame_step = frame_size*sample_rate, frame_stride*sample_rate
signal_length = len(emphasized_signal)
frame_length = int(round(frame_length))
num_frames = int(np.ceil(float(np.abs(signal_length - frame_length)) / frame_step))
pad_signal_length = num_frames * frame_step + frame_length
z = np.zeros(int(pad_signal_length - signal_length))
pad_signal = np.append(emphasized_signal, z)
indices = np.tile(np.arange(0, frame_length), (num_frames, 1)) + np.tile(np.arange(0, num_frames*frame_step, frame_step), (frame_length, 1)).T
frames = pad_signal[indices.astype(np.int32, copy=False)]
ham = np.hamming(frame_length)
# plt.plot(ham)
# plt.show()

frames *= ham

NFFT = 512
mag_frames = np.absolute(np.fft.rfft(frames, NFFT))
pow_frames = ((1.0 / NFFT) * ((mag_frames) ** 2))  #(348, 257)

nfilt = 40
low_freq_mel = 0
high_freq_mel = (2595 * np.log10(1 + (sample_rate/2) / 700))

mel_points = np.linspace(low_freq_mel, high_freq_mel, nfilt + 2)

hz_points = (700 * (10**(mel_points / 2595) - 1))

bin = np.floor((NFFT + 1)*hz_points / sample_rate)
fbank = np.zeros((nfilt, int(np.floor(NFFT / 2 + 1))))
for m in range(1, nfilt + 1):
    f_m_minus = int(bin[m-1])
    f_m = int(bin[m])
    f_m_plus = int(bin[m+1])

    for k in range(f_m_minus, f_m):
        fbank[m-1, k] = (k-bin[m-1]) / (bin[m]-bin[m - 1])
    for k in range(f_m, f_m_plus):
        fbank[m-1, k] = (bin[m + 1] - k) / (bin[m + 1] - bin[m])

filter_banks = np.dot(pow_frames, fbank.T)
filter_banks = np.where(filter_banks == 0, np.finfo(float).eps, filter_banks)
filter_banks = 20 * np.log10(filter_banks)


plt.imshow(np.flipud(filter_banks.T), cmap=plt.cm.jet, aspect=0.2, extent=[0,filter_banks.shape[1],0,filter_banks.shape[0]],shape=[11, 9])
plt.axis("normal")
plt.savefig('./test2.png')
plt.show()

运行的结果如下：