基于分类特征空间高斯混合模型和神经网络融合的说话人识别
Speaker Identification Based on Classify Feature Sub-space Gaussian Mixture Model and Neural Net Fusion
-
摘要: 该文提出了一种基于分类高斯混合模型和神经网络融合(FS-GMM/NN)的说话人识别方法，通过对特征矢量进行聚类分析，将说话人的训练语音分成若干类。然后根据各个类中含特征矢量的多少采用不同的模型混合度，训练建立分类高斯混合模型。并采用神经网络实现各个分类高斯混合模型输出的融合。在100个男性话者的与文本无关的说话人识别实验中，基于分类高斯混合模型和神经网络融合的方法在识别性能及噪声鲁棒性上都优于不分类的GMM识别系统，并具有较高的模型训练效率，且可以有效地降低话者模型的混合度和测试语音长度。Abstract: In this paper, a speaker identification system is proposed based on classify Fea-ture Sub-space Gaussian Mixture Model and Neural Net fusion (FS-GMM/NN) . With clus-tering analysis of the feature vectors, the speakers training feature vectors can be classified to some subsets and training classify Gaussian Mixture Models (GMM) with different mix-tures according to the subsets feature vectorss number. Finally, the outputs of every classify GMM will be fused by Neural Net (NN). In the experiment of text-independent speaker iden-tification of 100 speakers (male), the system based on FS-GMM/NN overmatch the Baseline Gaussian Mixture Model (B-GMM) in identification performance and noise robustness with fewer mixtures and shorter test speech. Moreover, the training of FS-GMM/NN is more effective.
-
Reynolds D A, Rose R C. Robust text-independent speaker identification using Gaussian mixture speaker models[J].IEEE Trans. on Speech Audio Process.1995, 3(1):72-83[2]Reynolds D A. Speaker identification and verification using Gaussian mixture speaker models[J].Speech Communication.1995, 17(1-2):91-108[3]Reynolds D A. Speaker verification using adapted Gaussian mixture models[J].Digital Signal Processing.2000, 10(1-3):19-41[4]Deller J R, Proakisa J G, Hansenm J H L. Discrete-Time Processing of Speech Signals. New York: Macmillan Publishing Company, 1993.[5]Reynolds D A. Experimental evaluation of features for robust speaker identification[J].IEEE Trans.on Speech Audio Process.1994, 2(4):639-643[6]Chang E, Shi Y, Zhou J, Huang C. Speech lab in a box: A mandarin speech toolbox to jumpstart speech related research. in EUROSPEECH, Aalborg, Denmark, 2001: 192-199. -
计量
- 文章访问数: 2756
- HTML全文浏览量: 142
- PDF下载量: 1057
- 被引次数: 0
下载:
