CAJ | 학술논문

为了解决基于相位差滤波器（PBF）双麦克风方法残留噪声较多的问题，本文在 PBF 方法基础上提出一种基于高斯混合模型的双麦克风噪声消除方法。该方法首先采用高斯混合模型（GMM）对目标语音存在（λ1）与目标语音不存（λ0）在这两种情况进行建模。其次，在实时增强阶段，根据贝叶斯分类器计算每帧的目标语音存在概率（TSPP），随后根据噪声抑制最大化准则修正 PBF 的增益函数并得到改进的相位差滤波器（IPBF），最后将TSPP 与 IPBF 的增益函数相结合，进而得到一种用于双麦克风噪声消除的掩蔽滤波器。实验结果表明：本文提出算法可有效抑制残留噪声，尤其是在目标语音不存在的时间段。
위료해결기우상위차려파기（PBF）쌍맥극풍방법잔류조성교다적문제，본문재 PBF 방법기출상제출일충기우고사혼합모형적쌍맥극풍조성소제방법。해방법수선채용고사혼합모형（GMM）대목표어음존재（λ1）여목표어음불존（λ0）재저량충정황진행건모。기차，재실시증강계단，근거패협사분류기계산매정적목표어음존재개솔（TSPP），수후근거조성억제최대화준칙수정 PBF 적증익함수병득도개진적상위차려파기（IPBF），최후장TSPP 여 IPBF 적증익함수상결합，진이득도일충용우쌍맥극풍조성소제적엄폐려파기。실험결과표명：본문제출산법가유효억제잔류조성，우기시재목표어음불존재적시간단。
In order to reduce the residual noise caused by the method of phase-error based filter,a noise reduction ap-proach based on dual microphones is proposed based on the Gaussian Mixture Model (GMM).First,two cases,i.e.target speech presence and target speech absence were modeled by the GMM,respectively.Then,the frame-based target speech present probability (TSPP)is calculated with the Bayesian classification.Next,an improved PBF (IPBF)is presented by modifying the gain function (GF)of PBF using the rule of noise suppression maximization.Finally,a mask filter for noise reduction is obtained by combining the GF of IPBF and TSPP.Some simulation results reveal that the proposed method could suppress the residual noise effectively,particularly in the period of target speech absence.