结合差分阵列与幅度谱减的双麦语音增强算法

为了抑制小型语音通信设备中的方向性噪声干扰问题，提出了一种结合差分阵列与幅度谱减的双麦语音增强算法。该算法首先利用一阶差分阵列技术，对两麦克风采集到的带噪语音信号进行处理，得到语音通道信号和噪声通道信号。接着利用差分阵列处理后的两通道信号对语音通道信号的信噪比进行估计。最后利用幅度谱减法对语音通道信号中残留噪声进行消除。针对语音通道信号的信噪比估计，本文给出了两种新奇的计算方法。仿真实验表明，该算法有效的抑制了方向噪声，改善了语音的质量，去噪效果及语音质量均优于对比算法。 

Dual-Microphone Speech Enhancement Algorithm Combined with Differential Microphone Arrays and Amplitude Spectral subtraction

To suppress the directional noise in miniaturized speech devices, a dual - microphone speech enhancement algorithm combined with differential arrays and amplitude spectrum reduction is proposed. First, the speech channel and noise channel signals were obtained by using the first-order differential arrays to process the noisy speech collected by two microphones. Then, the signal-to-noise ratio of the speech channel signals is estimated using the two channel signals processed by the differential arrays. Finally, the residual noise in speech channel signal is eliminated using the amplitude spectrum subtraction. In this paper, two novel methods are proposed to estimate the signal-to-noise ratio of speech channel signals. The simulation results show that the algorithm is effective in suppressing the directional noise, improving the quality of speech, the denoising effect and speech quality are better than the comparison algorithm.