To Weight or Not to Weight: Source-Normalised LDA for Speaker Recognition Using i-vectors

@inproceedings{McLaren2011ToWO,
  title={To Weight or Not to Weight: Source-Normalised LDA for Speaker Recognition Using i-vectors},
  author={Mitchell McLaren and David A. van Leeuwen},
  booktitle={INTERSPEECH},
  year={2011}
}
Source-normalised Linear Discriminant Analysis (SNLDA) was recently introduced to improve speaker recognition using i-vectors extracted from multiple speech sources. SNLDA normalises for the effect of speech source in the calculation of the between-speaker covariance matrix. Sourcenormalised-and-weighted (SNAW) LDA computes a weighted average of source-normalised covariance matrices to better exploit available information. This paper investigates the statistical significance of performance… CONTINUE READING
Highly Cited
This paper has 44 citations. REVIEW CITATIONS

Citations

Publications citing this paper.
Showing 1-10 of 23 extracted citations

Similar Papers

Loading similar papers…