语音识别

'''
    语音识别：声音的本质是震动，震动的本质是位移关于时间的函数，波形文件(.wav)中记录了不同采样时刻的位移。
            通过傅里叶变换，可以将时间域的声音函数分解为一系列不同频率的正弦函数的叠加，通过频率谱线的特殊分布，
            建立音频内容和文本的对应关系，以此作为模型训练的基础。

            梅尔频率倒谱系数(MFCC)通过与声音内容密切相关的13个特殊频率所对应的能量分布，可以使用梅尔频率倒谱系数矩阵作为语音识别的特征。
            基于隐形马尔科夫模型进行模式识别，找到测试样本最匹配的声音模型，从而识别语音内容。

    梅尔频率倒谱系数相关API：
            import scipy.io.wavfile as wf
            import python_speech_features as sf

            sample_rate, sigs = wf.read('../data/freq.wav')
            mfcc = sf.mfcc(sigs, sample_rate)

'''
import scipy.io.wavfile as wf  # 读取 音频文件
import python_speech_features as sf  # 提取MFCC
import numpy as np

import matplotlib.pyplot as mp

sample_rate, sigs = wf.read('./ml_data/speeches/training/apple/apple01.wav')
print(sample_rate, sigs.shape)
# 获取MFCC矩阵
mfcc = sf.mfcc(sigs, sample_rate)
print(mfcc.shape)

mp.matshow(mfcc.T, cmap='gist_rainbow')
mp.show()


输出结果：
8000 (2694,)
(33, 13)

 

'''
    隐马尔科夫模型相关API：----音频识别
            import hmmlearn.hmm as hl
            # n_components: 用几个高斯分布函数拟合样本数据
            # covariance_type: 相关矩阵的辅对角线进行相关性比较
            # n_iter: 最大迭代上限
            model = hl.GaussianHMM(n_components=4, covariance_type='diag', n_iter=1000)
            model.fit(mfccs)
            # 使用模型匹配测试mfcc矩阵的分值
            score = model.score(test_mfccs)

    案例：对speeches文件夹下的七个单词文件进行语音识别
        步骤：
            1.读取每个音频类别的文件夹，加载每个类别的音频文件数据
            2.得到每个音频的mfcc，训练HMM模型
            3.每个类别将会训练得到一个HMM模型，并保存模型
            4.读取测试数据，得到测试音频的mfcc
            5.使用每个HMM模型对该mfcc进行得分预测，取其大者作为音频类别
'''
import os
import numpy as np
import scipy.io.wavfile as wf
import python_speech_features as sf
import hmmlearn.hmm as hl


def search_files(directory):
    # 读取目录的内容，返回一个字典：{'apple':[url1,url2...],'banana':[url1,url2...],}
    # 把directory目录改为当前平台系统所能识别的目录
    directory = os.path.normpath(directory)
    objects = {}
    # curdir当前文件夹，subdirs子文件夹，files子文件夹项下的文件
    for curdir, subdirs, files in os.walk(directory):
        for file in files:
            if file.endswith('.wav'):
                label = curdir.split(os.path.sep)[-1]
                if label not in objects:
                    objects[label] = []
                path = os.path.join(curdir, file)
                objects[label].append(path)
    return objects


print(search_files('../ml_data/speeches/training/'))

# 整理训练集
train_samples = search_files('../ml_data/speeches/training/')
train_x, train_y = [], []
for label, filenames in train_samples.items():
    mfccs = np.array([])
    for filename in filenames:
        sample_rate, sigs = wf.read(filename)
        mfcc = sf.mfcc(sigs, sample_rate)
        if len(mfccs) == 0:
            mfccs = mfcc
        else:
            mfccs = np.append(mfccs, mfcc, axis=0)
    train_x.append(mfccs)
    train_y.append(label)

# 基于HMM模型，训练样本
# models: {'apple':<model object>, ...}
models = {}
for mfccs, label in zip(train_x, train_y):
    model = hl.GaussianHMMz(n_components=4, covariance_type='diag', n_iter=1000)
    models[label] = model.fit(mfccs)

# 针对测试集样本进行测试

# 整理测试集
test_samples = search_files('../ml_data/speeches/testing/')
test_x, test_y = [], []
for label, filenames in test_samples.items():
    mfccs = np.array([])
    for filename in filenames:
        sample_rate, sigs = wf.read(filename)
        mfcc = sf.mfcc(sigs, sample_rate)
        if len(mfccs) == 0:
            mfccs = mfcc
        else:
            mfccs = np.append(mfccs, mfcc, axis=0)
    test_x.append(mfccs)
    test_y.append(label)

# test_x与test_y有7组数据
# 遍历test_x,使用每个HMM模型对同一个样本进行得分判比
pred_test_y = []
for mfccs in test_x:
    best_score, best_label = None, None
    for label, model in models.items():
        score = model.score(mfccs)
        if (best_score is None) or (best_score < score):
            best_score = score
            best_label = label
    pred_test_y.append(best_label)

print(test_y)
print(pred_test_y)