摘要
提出一种利用人耳听觉特性的双声道回波抵消方法,与传统方法相比能够更快地收敛到“真实”回波路径.我们用子带分解缩短自适应滤波器长度,减小算法复杂度,并在子带分解基础上,利用人耳听觉特性降低输入信号的相关度,使自适应滤波收敛更快.数值实验表明了该方法的有效性.
In this paper, a sub-band stereo acoustic echo cancellation using human auditory characteristics is proposed. This method can converge to ＂true＂ echo path more quickly than traditional methods. The length of adaptive filters is shortened, and processing complexity reduced by using sub-band decomposition. Based on the sub-band decomposition, correlation of input signals can be reduced by using human auditory characteristics. Thus, the adaptive filters converge faster. Simulation results show validity of the proposed method.
出处
《应用科学学报》
CAS
CSCD
北大核心
2007年第4期331-336,共6页
Journal of Applied Sciences
基金
国家自然科学基金(60472058)
教育部博士点基金(20050286001)
教育部新世纪优秀人才支持计划(NCET-04-0483)资助项目
关键词
回波抵消
人耳听觉特性
子带分解
acoustic echo cancellation （AEC）
human auditory characteristics
sub-band decomposition
