Beamforming Using Uniform Circular Arrays for Distant Speech Recognition in Reverberant Environments and Double-Talk Scenarios

摘要

Beamforming is crucial for hands-free mobile terminals and voice-enabled automated home environments based on distant-speech interaction to mitigate causes of system degradation, e.g., interfering noise sources or competing speakers. This paper presents an adaptation of the most common state-of-the-art broadband beamformers to uniform circular arrays, such that competing speakers are attenuated sufficiently for distant speech recognition. As a result, a new beamformer is presented. Finally, the speech quality of the beamformers' enhanced signals is evaluated with different objective speech quality measures and a word recognizer as a measure for the attenuation of competing speakers.