一种改进的广义旁瓣对消阵列语音增强算法  

An Improved Speech Enhancement Algorithm for GSC Microphone

作　　者：童环 夏秀渝[1] TONG Huan;XIA Xiuyu(School of Electronic Information,Sichuan University,Chengdu 610064,China)

机构地区：[1]四川大学电子信息学院,四川成都610065

出　　处：《成都信息工程大学学报》2024年第3期263-267,共5页Journal of Chengdu University of Information Technology

摘　　要：传统广义旁瓣对消(generalized sidelobe cancellation,GSC)算法在复杂声学环境下性能较差,语音增强效果不理想,需要加强其去噪能力以提高输出语音质量。对此,提出一种改进型广义旁瓣对消算法。构建基于频域GSC和时域GSC两级滤波的系统结构,利用一个选择滤波器输出两级滤波中质量较好的语音;并将GSC上支路权值修改为可自适应调节的形式,提高算法的适应性;GSC权值迭代时采用一种变步长自适应算法,步长因子根据信号信干噪比实时调整,防止滤波器权值发散。实验结果表明,相比于传统GSC算法,新算法在SINR、PESQ、STOI和SDR等指标上都有提升。The traditional Generalized Sidelobe Cancellation(GSC)algorithm has poor performance in complex acoustic environments,and the effect of speech enhancement is not ideal.In this regard,an improved GSC algorithm is proposed.In this paper,a system structure based on frequency-domain GSC and time-domain GSC two-stage filtering is constructed,and a selected filter is used to output speech with better quality;Adaptability;and a variable step size adaptive algorithm is proposed,and the step size factor is adjusted in real time according to the Signal to Interference plus Noise Ratio(SINR)to prevent the divergence of the filter weight.The experimental results show that compared with the traditional GSC algorithm,the algorithm in this paper has improved in terms of SINR、PESQ、STOI,and SDR.

关 键 词：麦克风阵列 语音增强 GSC 变步长自适应算法 

分 类 号：TN912.35[电子电信—通信与信息系统]