A comparison of neural network feature transforms for speaker diarization

@inproceedings{Yella2015ACO,
  title={A comparison of neural network feature transforms for speaker diarization},
  author={Sree Harsha Yella and Andreas Stolcke},
  booktitle={INTERSPEECH},
  year={2015}
}
Speaker diarization finds contiguous speaker segments in an audio stream and clusters them by speaker identity, without us ing a-priori knowledge about the number of speakers or enrollme nt data. Diarization typically clusters speech segments base d on short-term spectral features. In prior work, we showed that neural networks can serve as discriminative feature transform ers for diarization by training them to perform same/different spe aker comparisons on speech segments, yielding improved… CONTINUE READING

References

Publications referenced by this paper.
Showing 1-10 of 19 references

W

  • I. Mccowan, G. Lathoud, M. Lincoln, A. Lisowska
  • Post , D. Reidsma, and P. Wellner, “The AMI…
  • 2005
Highly Influential
11 Excerpts

Similar Papers

Loading similar papers…