ICASSP
2011
IEEE
13 years 3 months ago
2011
IEEE
Model-based methods for sequential organization in cochannel speech require pretrained speaker models and often prior knowledge of participating speakers. We propose an unsupervis...