Purity Algorithms for Speaker Diarization of Meetings Data

  title={Purity Algorithms for Speaker Diarization of Meetings Data},
  author={Xavier Anguera Mir{\'o} and Chuck Wooters and Javier Hernando},
  journal={2006 IEEE International Conference on Acoustics Speech and Signal Processing Proceedings},
When performing speaker diarization, it is common to use an agglomerative clustering approach where the acoustic data is first split in small pieces and then pairs are merged until reaching a stopping point. When using a purely agglomerative clustering technique, one cluster cannot be split into two. Therefore, errors caused by multiple speakers being assigned to one cluster can be common. Furthermore, clusters often contain non-speech frames, creating problems when deciding which two clusters… CONTINUE READING
Highly Cited
This paper has 34 citations. REVIEW CITATIONS