【摘要】 目的:研究数字助听器中竞争语音情形下的语音增强算法。方法:针对语音增强算法中的麦克风阵列自适应波束形成算法,简要回顾了广义旁瓣消除器(GSC)结构的鲁棒性时域自适应波束形成算法,研究了频域GSC结构的自适应波束形成算法。结果:设计仿真实验,对比时、频域GSC的计算复杂度和收敛速度。结论:仿真结果显示,频域GSC结构的自适应波束形成算法与时域GSC对比,具有计算速度快、收敛性能好的优点。更多还原

【Abstract】 Objective To deal with the digital hearing aids in competing talker situations using speech enhancement algorithm.Methods For adaptive beam forming algorithm using microphone array,the frequency-domain adaptive beamforming algorithm based on robustness time-domain GSC structure was studied.Results The simulation was designed for the frequency-domain adaptive beamforming algorithm,and the computational complexity and convergence rate were compared between the frequency domain GSC and the time domain GSC.Conclusion The simulation results show that the frequency-domain GSC adaptive beamforming algorithm has a higher calculation speed and convergence rate,and thus has more research significance when compared with time-domain GSC.更多还原