基于TensorFlow的歌曲曲风变换

背景：

在图像上的风格变换（Style Transfer,论文，基于Torch的代码实现：neural-style）同样可以应用于音频中。
本文采用的是英文TravelingLight.mp3和东风破.mp3作为音频源和参考源。考虑到内存有限，仅仅截取音频中的10s进行风格的变换。

代码：

# -*- coding: utf-8 -*-__author__ = 'jason'import tensorflow as tfimport librosa# 用来提取音频文件, 参看<中文语音识别>import numpy as npimport osimport pdb#import shlex  # python2 pipes# 音频文件路径content_audio = "TravelingLight.mp3"style_audio = "东风破.mp3"# 为英文歌曲<Traveling Light>添加周杰伦风味# 剪辑一段音频, 默认取开头的10s, 太大内存吃不消def cut_audio(filename, start_pos='00:00:00', lens=10):    newfile = os.path.splitext(os.path.basename(filename))[0] + '_' + str(lens) + 's.mp3'    # 确保系统中已安装ffmpeg,这是ffmpeg的命令行方式,更加详细的使用方法可以google    cmd = "ffmpeg -i {} -ss {} -t {} -acodec copy {}".format(filename, start_pos, lens, newfile)    os.system(cmd)    return newfile#上面的ffmpeg注意-acodec copy参数，否则会报错content_audio_10s = cut_audio(content_audio, start_pos='00:00:33')style_audio_10s = cut_audio(style_audio, start_pos='00:00:38')#content_audio_10s = "TravelingLight_10s.mp3"#style_audio_10s = "东风破_10s.mp3" # Short Time Fourier Transform音频转spectrogram（把1维信号转为2维, 可以被视作图像）# https://en.wikipedia.org/wiki/Short-time_Fourier_transformN_FFT = 2048def read_audio(filename):    x, fs = librosa.load(filename)    S = librosa.stft(x, N_FFT)    p = np.angle(S)    S = np.log1p(np.abs(S[:,:430]))    return S, fscontent_data, _ = read_audio(content_audio_10s)style_data, fs = read_audio(style_audio_10s)samples_n = content_data.shape[1]  # 430channels_n = style_data.shape[0]   # 1025style_data = style_data[:channels_n, :samples_n]content_data_tf = np.ascontiguousarray(content_data.T[None,None,:,:])style_data_tf = np.ascontiguousarray(style_data.T[None,None,:,:])# filter shape "[filter_height, filter_width, in_channels, out_channels]"N_FILTERS = 4096std = np.sqrt(2) * np.sqrt(2.0 / ((channels_n + N_FILTERS) * 11))kernel = np.random.randn(1, 11, channels_n, N_FILTERS)*std# content and style featuresg = tf.Graph()with g.as_default(), g.device('/cpu:0'), tf.Session() as sess:    # data shape "[batch, in_height, in_width, in_channels]",    x = tf.placeholder('float32', [1, 1, samples_n, channels_n], name="x")    kernel_tf = tf.constant(kernel, name="kernel", dtype='float32')    conv = tf.nn.conv2d(x, kernel_tf, strides=[1, 1, 1, 1], padding="VALID", name="conv")    net = tf.nn.relu(conv)    content_features = net.eval(feed_dict={x: content_data_tf})    style_features = net.eval(feed_dict={x: style_data_tf})    features = np.reshape(style_features, (-1, N_FILTERS))    style_gram = np.matmul(features.T, features) / samples_n# OptimizeALPHA= 0.01   # ALPHA越大,content越占主导; 如果ALPHA为0,表示没有contentresult = Nonewith tf.Graph().as_default():    learning_rate= 0.001    x = tf.Variable(np.random.randn(1, 1, samples_n, channels_n).astype(np.float32)*learning_rate, name="x")    kernel_tf = tf.constant(kernel, name="kernel", dtype='float32')    conv = tf.nn.conv2d(x, kernel_tf, strides=[1, 1, 1, 1], padding="VALID", name="conv")    net = tf.nn.relu(conv)    content_loss = ALPHA * 2 * tf.nn.l2_loss(net - content_features)    style_loss = 0    _, height, width, number = map(lambda i: i.value, net.get_shape())    size = height * width * number    feats = tf.reshape(net, (-1, number))    gram = tf.matmul(tf.transpose(feats), feats)  / samples_n    style_loss = 2 * tf.nn.l2_loss(gram - style_gram)    # loss    loss = content_loss + style_loss    opt = tf.contrib.opt.ScipyOptimizerInterface(loss, method='L-BFGS-B', options={'maxiter': 300})    # Optimization    init_op = tf.initialize_all_variables()#注意这里的初始化操作    with tf.Session() as sess:        #sess.run(tf.global_variables_initializer())#这是原来的初始化操作,会报错        sess.run(init_op)        opt.minimize(sess)        result = x.eval()# 把spectrogram转回wav音频audio = np.zeros_like(content_data)audio[:channels_n,:] = np.exp(result[0,0].T) - 1p = 2 * np.pi * np.random.random_sample(audio.shape) - np.pifor i in range(500):    S = audio * np.exp(1j*p)    x = librosa.istft(S)    p = np.angle(librosa.stft(x, N_FFT))librosa.output.write_wav("output.mp3", x, fs)

结果：

运行后的数据结果下载