基于GMM-UBM的飞机发动机声音识别方法研究

Research of Aircraft Engine Sound Recognition Method Based on GMM-UBM

摘要: 高斯混合模型–通用背景模型(Gaussian mixture model-universal background model, GMM-UBM)是说话人识别技术中最为常用的模型，该模型在诸多试验中都取得了很好的效果。本设计探索把GMM-UBM模型用在异常声音检测中，通过对飞机发动机声音信号的处理，提取梅尔频率倒谱(MFCC)特征参数，训练UBM模型，用MAP自适应的算法得到GMM-UBM模型，用GMM-UBM模型检测识别发动机声音。实验证明，该方法优化了由于外界干扰变化导致的识别率下降的问题。

Abstract: Gaussian mixture model-universal background model (GMM-UBM) is the most commonly used model in speaker recognition technology; the model has achieved very good results in many ex-periments. In this design, the GMM-UBM model is used in the abnormal sound detection. First, we process the aircraft engine sound signal, second extract the MFCC characteristic parameters, then train UBM model and last obtain the GMM-UBM model by MAP adaptive algorithm. The ultimate goal of the test indicates that the method could optimize the recognition rate decline due to interference change.

文章引用：杨毫鸽, 孙成立. 基于GMM-UBM的飞机发动机声音识别方法研究[J]. 计算机科学与应用, 2017, 7(8): 781-787. https://doi.org/10.12677/CSA.2017.78089

参考文献

[1]	Xie, C., Cao, X.L. and He, L.L. (2012) Algorithm of Abnormal Audio Recogniton Based on Improved MFCC. Procedia Engineering, 29, 731-737. https://doi.org/10.1016/j.proeng.2012.01.032
[2]	张正平. 基于GMM-UBM说话人模型的连续自适应算法研究[J]. 通信电源技术, 2016, 33(2).
[3]	Wu, J.Q. and Yu, J.J. (2011) A Improved Arithmetic of MFCC in Speech Recognition System. IEEE Electronics, Communication and Control (ICECC), 719-722.
[4]	Harma, A., McKinney, M.F. and Skowronek, J. (2005) Automatic Surveillance of the Acoustic Activity in Our Living Environment. Proc of IEEE ICME, 634-637.
[5]	赵力. 语音信号处理[M]. 北京: 机械工业出版社, 2012: 75-134.
[6]	李燕萍. 说话人辨认中的特征参数提取和鲁棒性技术研究[D]: [硕士学位论文]. 南京: 南京理工大学, 2009.
[7]	王炳锡, 屈丹, 彭煊. 实用语音识别基础[M]. 北京: 国防工业出版社, 2005.
[8]	王伟, 邓辉文. 基于MFCC参数和VQ的说话人识别系统[J]. 仪器仪表学报, 2006, 27(6): 2253-2155.
[9]	Ito, A. and Aiba, A. (2009) Dectetion of Abnormal Sound Using Multi-Stage GMM for Surveillance Microphone. International Conference on Information Assuranceand Security.
[10]	Mark, M.-W. and Rao, W. (2011) Utterance Partitioning with Caoustic Vector Resampling for GMM-SVM Speaker Verification. Speech Communication, 53, 119-130. https://doi.org/10.1016/j.specom.2010.06.011
[11]	Ntalampiras, S., Potamitis, I. and Fakotakis, N. (2009) An Adaptive Framework for Acoustic Monitoring of Potential Hazards. EURASIP Journal on Audio, Speech, and Music Processing, 2, 1-15.