Using Energy Difference for Speech Separation of Dual - microphone Close - talk System

Using the computational auditory scene analysis (CASA) as a framework, a novel speech separation approach based on dual-microphone energy difference (DMED) is proposed for close-talk system. The energy levels of the two microphones are calculated in time-frequency (T-F) units. The DMEDs are calculated as the energy level ratio between the two microphones, and used as a cue to estimate the signal to noise ratio (SNR) and ideal binary mask (IBM) for mix-acoustic of the close-to-mouth microphone. The binary masked units are grouped to generate the target speech. Test with speeches and different noises show that the algorithm is more than 95 % accurate. As the T-F units’ length increase, the accuracy increase as well. Using automatic speech recognition (ASR) analysis, we show that the proposed algorithm improves speech quality in actual close talk system. Copyright © 2013 IFSA.