Study on Multichannel Speech Enhancement Technology in Voice Human-Computer Interaction

As the primary research area of the Multimoda1 Human-computer Interaction, Voice Interaction mainly involves extraction and identification of the natural speech signal, where the former provides the reliable signal sources, which are analyzed by the latter. The multichannel speech enhancement technology is studied in this paper, aiming at the Voice Interactive. The simulated results show the effectiveness and superiority of the improved algorithm proposed in the paper.