Cochannel Talker Interference Suppression

Cochannel talker interference suppression is defined as the processing of a waveform containing two simultaneous speech signals, referred to as the target and the jammer, to produce a signal containing an estimate of the target speech signal alone. The first part of this report describes the evaluation of a simulated suppression system that attenuates the jammer component of a cochannel signal, given the voicing states (voiced, unvoiced, silent) of the target and jammer speech as a function of time and given the isolated target and jammer speech waveforms. Ten listeners heard sentence pairs at average target-to-jammer ratios from -3 to -15 dB. Generally, 10 to 20 dB of jammer attenuation during regions of voiced target or jammer improved target intelligibility, but the level of improvement was speaker-dependent. These results are important because they upper-bound the performance of earlier systems operating only in the voiced talker regions. The second part addresses the problem of speaker activity detection.