Loading...
Please wait, while we are loading the content...
Similar Documents
Blind speech separation in a meeting situation with maximum snr beamformers.
| Content Provider | CiteSeerX |
|---|---|
| Author | Araki, Shoko Sawada, Hiroshi Makino, Shoji |
| Abstract | We propose a speech separation method for a meeting situation, where each speaker sometimes speaks and the number of speakers changes every moment. Many source separation methods have al-ready been proposed, however, they consider a case where all the speakers keep speaking: this is not always true in a real meeting. In such cases, in addition to separation, speech detection and the classification of the detected speech according to speaker become important issues. For that purpose, we propose a method that em-ploys a maximum signal-to-noise (MaxSNR) beamformer combined with a voice activity detector and online clustering. We also discuss the scaling ambiguity problem as regards the MaxSNR beamformer, and provide their solutions. We report some encouraging results for a real meeting in a room with a reverberation time of about 350 ms. Index Terms — Speech separation, maximum SNR beamformer, scaling ambiguity, voice activity detector, online clustering. 1. |
| File Format | |
| Access Restriction | Open |
| Subject Keyword | Real Meeting Voice Activity Detector Online Clustering Index Term Speech Separation Speech Detection Speech Separation Method Encouraging Result Reverberation Time Maxsnr Beamformer Maximum Snr Beamformer Important Issue Many Source Separation Method Scaling Ambiguity Problem Meeting Situation Maximum Signal-to-noise Speaker Change Detected Speech |
| Content Type | Text |
