摘要： 提出一种符合人耳听觉感知的语音增强方法，使电子耳蜗能在噪声环境下获得准确的语音信息。利用Bark子波变换实现电子耳蜗中的语音处理，结合人耳听觉系统特性实现语音增强。使用根据人耳听觉掩蔽效应提出的自适应减参数。实验结果表明该算法在低信噪比情况下，信噪比可提高30 dB左右，更好地抑制了残留噪声和背景噪声，合成的语音具有较好清晰度和可懂度。
关键词:
电子耳蜗,
语音增强,
掩蔽效应,
Bark子波
Abstract: To acquire accurate speech information in cochlear implant under noise enviroments, a speech enhancement in cochlear implant based on human auditory masking properties is proposed. Bark wavelet transform is used to realize speech in electronic cochlear implant and to realize speech enhancement combined with human ear’s auditory system. Subtract parameter is got based on the masking properties of human auditory and self-adaptive.The result shows that the SNR output can be improved by 30 dB even if the SNR input is low and can reduce the residual noise and backgrand noise more effectively. The enhanced speech is more clear and intelligible.
Key words:
electronic cochlear implant,
speech enhancement,
auditory masking properties,
Bark wavelet
中图分类号:
黄雅婷;陶 智;顾济华;赵鹤鸣;严冬明. 基于人耳掩蔽效应的电子耳蜗语音增强方法[J]. 计算机工程, 2008, 34(10): 280-282.
HUANG Ya-ting; TAO Zhi; GU Ji-hua; ZHAO He-ming; YAN Dong-ming. Speech Enhancement Method in Electronic Cochlear Implant Based on Human Auditory Masking Properties[J]. Computer Engineering, 2008, 34(10): 280-282.