基于DTW算法语音识别系统的仿真及DSP实现  

Speech Recognition System Simulation Based on DTW and DSP Implementation

作　　者：陈锡锻 王瑞[1] 肖雄[1] 洪涛[1] 

机构地区：[1]浙江工业大学信息工程学院,浙江杭州310014

出　　处：《电声技术》2013年第12期66-69,共4页Audio Engineering

摘　　要：DTW(Dynamic Time Warping)算法的实现简单有效,在孤立词语音识别系统中得到了广泛的应用。采用谱减法进行前端去噪处理,利用Matlab对语音识别系统进行了仿真,并设计了一种以16位数字信号处理器TMS320VC5509为核心的孤立词语音识别系统。实验结果表明,系统能满足实时性能要求,识别效果良好。The realization of the DTW algorithm is simple and effective, so it is widely used in the isolated-word speech recognition system. In this paper, the spectral subtraction is used in front end denoising processing, the simulation is carried on by using Matlab, and a kind of isolated word speech recognition system is designed which is based on the 16 digital signal processor named TMS320VC5509. The simulation results show that the system can meet the real-time perforence require- ments and provide a good performance in the speech recognition.

关 键 词：DTW 谱减法 DSP 音频芯片 

分 类 号：TN912[电子电信—通信与信息系统]