摘要
首先讨论了回声消除中的双端发声问题,指出传统的双端发声检测方法用于回声消除时,其不可避免的误检会导致回声消除系统性能严重恶化,使系统不稳定。本文提出一种使用双滤波器的回声消除算法。该算法使用自适应滤波器跟踪回声信道,使用一个辅助滤波器和自适应滤波器一起完成回声消除。实验结果表明,该算法与传统算法相比,运算量相当;但在双端发声期间,该算法稳定性更高,回音往返损耗增强(E cho return loss en-hancem en t,ERLE)有明显的改善。
The double-talk problem in the echo cancellation is discussed. It is pointed out that the false detection results in the echo cancellation system to be unstable when traditional dou- ble-talk detection methods are applied in the echo cancellatign. So, a new algorithm with two- filters for the echo cancellation is proposed. The algorithm uses an adaptive filter to track the echo path, and an auxiliary filter to cancel the echo together with the adaptive filter. Experimental results show that, compared with traditional double-talk detection methods, the new algorithm has comparable computation complexity and works more stably, thus improving the echo return loss enhancement (ERLE) during the double-talk.
出处
《数据采集与处理》
CSCD
北大核心
2006年第1期74-79,共6页
Journal of Data Acquisition and Processing
基金
美国模拟器件公司大学研究计划资助项目
关键词
回声消除
自适应滤波
双端发声检测
双滤波器
echo cancellation
adaptive filtering
double-talk detection
two-filters