Single-channel speaker diarization based on spatial features
Mathieu Hu, Pablo Peso Parada, Dushyant Sharma, Simon Doclo, Toon van Waterschoot, Mike Brookes, Patrick A. Naylor · 2015
Speaker diarization has gained much importance over the past five years in helping overcome key challenges faced by automatic meeting transcription systems. Current state-of-the-art algorithms can only utilize spatial information when multi-microphone recordings are available. In this paper, we propose the novel use of reverberation as a source of spatial information obtained from single-channel recordings to perform speaker diarization. The proposed system is shown to reduce speaker classification errors by 34% when compared with current MFCC based single-channel systems.