IEEE 2016 International Conference on Wireless Communications, Signal Processing and Networking (WiSPNET 2016)

235 / 2016-01-21 20:28:00

Dual-channel speech separation by sub-segmental directional statistics

9209,9219

In this letter, we present a novel speech separation scheme using two microphones. We divide the inter-phase information into sub-segments and statistic these directional segments. Then we construct the objective function by convolving the statistics information with a low pass filter. By the decreasing gradient algorithm and ideal binary mask, we obtain the separated speeches. The method is valid for phase wrap-around aliasing problem which has the restriction that the microphones are closely spaced. Furthermore, the method can be applied to an underdetermined case where there are fewer microphones than sources. Our objective quality evaluation experiments demonstrate the effectiveness of the proposed method.