Denoising on Wavelet Compromise Threshold Algorithm for Robust Speech Recognition

摘要：

The performance degradation of speech recognition systems is a big problem in noisy environments. To obtain a high robust of speech recognition for noisy conditions, a new pre-processing stage based on wavelet denoising is proposed to extract robust features in the presence of additive white Gaussian noise. Recognition performance is compared with the commonly used Mel-Frequency Discrete Wavelet Coefficients (MFDWCs) with and without this preprocessing stage. The performance of speech signal with unscaled white noise could be highly improved in time domain.

关键词： Speech recognition wavelet transform compromise threshold

作者: Liu Xuefei Hu Chunhai

作者单位: Information Engineering Department,Environmental Management College of China,Qinhuangdao,066004 Chin Information Engineering Department,Environmental Management College of China,Qinhuangdao,066004China

会议类型: 国际会议

会议名称: The Second International Symposium on Test Automation & Instrumentation(ISTAI2008)(第二届国际自动化测试与仪器仪表学术会议)

会议地点: 北京

会议语种:英文

在线出版日期: 2008-11-01（万方平台首次上网日期，不代表论文的发表时间）