em算法matlab代码-HMM:唔

  • C8_960131
    了解作者
  • 231.2KB
    文件大小
  • zip
    文件格式
  • 0
    收藏次数
  • VIP专享
    资源类型
  • 0
    下载次数
  • 2022-05-19 08:09
    上传日期
em算法matlab代码HMM-单高斯v1.0 在这个项目中,我们想使用EM算法来处理针对孤立词数据的HMM训练。 还可以使用Viterbi算法考虑测试阶段。 结果表明,通过Matlab编程获得的性能与HTK相似。 在此项目中,我们尚未准备数据文件(.mfcc文件),请自行处理您自己的数据。 然后,您可能需要更改文件“ generate_trainingfile_list.m”和“ generate_testingfile_list.m”中的某些代码,直到数据文件路径为止。 请运行文件“ EM_HMM_isolated_digit_main.m”以启动。 有关更多信息,请发表评论。 您可以免费使用此作品。 !!! 更新时间:2017-09-07 您现在可以在网站上下载数据文件:请选择数据集:“隔离的TI数字培训文件,采样频率为8 kHz,终结点为:(isolated_digits_ti_train_endpt.zip)”,或者可以直接下载培训的.zip文件。仅通过此链接访问数据库: 训练数据: 。 测试数据: 请解压缩所有数据集,然后将训练和测试数据分别定位到目录“ wav \ iso
HMM-master.zip
  • HMM-master
  • wav2mfcc.m
    2KB
  • aij.mat
    14.2KB
  • var.mat
    14.2KB
  • wav2mfcc_e_d_a.m
    565B
  • generate_selected_TI_isolated_digits_testing_list_mat.m
    1.1KB
  • EM_HMM_FR.m
    4.5KB
  • EM_HMM_isolated_digit_main.m
    1.2KB
  • viterbi_dist_FR.m
    2.2KB
  • generate_selected_TI_isolated_digits_training_list_mat.m
    1KB
  • HMM.mat
    157.8KB
  • HMMtesting.m
    1.7KB
  • generate_trainingfile_list.m
    436B
  • obs.mat
    14.2KB
  • main_dr_wav2mfcc_e_d_a.m
    481B
  • generate_testingfile_list.m
    428B
  • README.md
    1.7KB
  • EM_HMMtraining.m
    5.5KB
  • mean.mat
    14.2KB
  • dr_wav2mfcc_e_d_a.m
    1.6KB
  • fwav2mfcc_e_d_a.m
    1.6KB
  • trainingfile_list.mat
    11KB
  • wav2logpow.m
    635B
  • slope.m
    761B
  • testingfile_list.mat
    11KB
内容介绍
# HMM-single-Gaussian-v1.0 In this project we would like to deal with training HMM for isolated words data applying EM algorithm. The testing phase is also considered using Viterbi algorithm. The results showed the performances which obtained by Matlab programming are similar to HTK's ones. In this project we have not prepared data files (.mfcc files) yet, please do it by yourself with your own data. Then, you may need to change some code in the file "generate_trainingfile_list.m" and "generate_testingfile_list.m" up to your data file paths. Please run the file "EM_HMM_isolated_digit_main.m" to start. For further information, please leave a comment. You may use this work free. !!! Update: 2017-09-07 You may now download data files on the website: http://www.ece.ucsb.edu/Faculty/Rabiner/ece259/speech%20recognition%20course.html Please select the data set: "isolated TI digits training files, 8 kHz sampled, endpointed: (isolated_digits_ti_train_endpt.zip)" to download it, or you may download directly the .zip file of training database only from this link: - training data: http://www.ece.ucsb.edu/Faculty/Rabiner/ece259/speech%20recognition%20course/databases/isolated_digits_ti_train_endpt.zip. - testing data: http://www.ece.ucsb.edu/Faculty/Rabiner/ece259/speech%20recognition%20course/databases/isolated_digits_ti_test_endpt.zip Please decompress all the data sets, then locate training and testing data into directories 'wav\isolated_digits_ti_train_endpt' and 'wav\isolated_digits_ti_test_endpt', respectively. Furthermore, we have just added some feature extracting functions that would help you to convert '.wav' files to '.mfc' files (feature files) Now you may run this project with only one click!
评论
    相关推荐