话筒
选择(遗传算法)
噪音(视频)
计算机科学
主动噪声控制
控制(管理)
声学
语音识别
人工智能
降噪
电信
物理
声压
图像(数学)
作者
Shota TOYOOKA,Yoshinobu KAJIKAWA
摘要
This paper proposes two methods for selecting a reference microphone from multiple microphones, focusing on the correlation between the reference signal and the desired signal and the causality constraint of the signal, in order to effectively reduce noise arriving from various locations. In the first method, a threshold is set for the correlation between the reference signal and the desired signal, and a microphone that satisfies the causality constraint is selected as the reference microphone from microphones with correlation values that exceed this threshold. On the other hand, the second method selects a microphone with the highest correlation between the reference signal and the desired signal as the reference microphone among the microphones that satisfy the causality constraint. We demonstrate through simulation experiments that the active noise control system using these proposed methods can provide stable and effective noise reduction for noise arriving from various locations. It is also shown that both methods can reselect an appropriate reference microphone when the noise source moves.
科研通智能强力驱动
Strongly Powered by AbleSci AI