Parameter optimization of the single channel late reverberation suppression technique

Refined recommendations for choosing optimal, in the sense of automatic speech recognition (ASR) accuracy maximum, parameters of the late reverberation suppression technique, have been proposed in this paper. It was shown that best value of boundary between early reflections and late reverberation approximates to 100 ms for ASR systems. It was shown also that, when estimating late reverberation power spectrum, an optimal value of averaging constant should be associated with statistical speech constants such as phoneme and stationarity durations. Several speech quality indicators were used also, and it was found that recognition accuracy is the best indicator in the sense of ability to inform the user about reached compromise between reverberation suppression and speech distortion.