摘要
对于可变噪声场,提出了一种自适应波束形成与后置滤波融合的波束形成麦克风阵列语音增强方法.针对自适应波束形成语音增强方法和后置滤波方法分别适用于强相干噪声场和非相干噪声场的现象,此方法将两种方法结合,既能抑制相干噪声,又能抑制非相干噪声.仿真实验表明,该方法对相干和非相干噪声场均有较好的消噪性能,因此具有较好的噪声鲁棒性.
For the variable noise field,a new microphone arrayspeech enhancement method combined adaptive beamfroming and post-filter is presented.The adaptive beamfroming method can only suppress coherent noises,and the post-filter method can only suppress incoherent noises.The new method can suppress both the coherent and the incoherent noises.Simulation results show that the method is effective in suppressing the coherent and the incoherent noises,which proves the robustness of the method for different noise fields.
出处
《大连交通大学学报》
CAS
2010年第4期29-32,共4页
Journal of Dalian Jiaotong University
基金
国家科技部创新基金资助项目(09C2622123243)
