Publication Details
Probabilistic embeddings for speaker diarization
Brummer Johan Nikolaas Langenhoven, Dr. (Phonexia)
Rohdin Johan A., Dr. (DCGM FIT BUT)
Stafylakis Themos (OMILIA)
Burget Lukáš, doc. Ing., Ph.D. (DCGM FIT BUT)
probabilistic embeddings, speaker diarization
Speaker embeddings (x-vectors) extracted from very short segments of speech have recently been shown to give competitive performance in speaker diarization. We generalize this recipe by extracting from each speech segment, in parallel with the x-vector, also a diagonal precision matrix, thus providing a path for the propagation of information about the quality of the speech segment into a PLDA scoring backend. These precisions quantify the uncertainty about what the values of the embeddings might have been if they had been extracted from high quality speech segments. The proposed probabilistic embeddings (x-vectors with precisions) are interfaced with the PLDA model by treating the x-vectors as hidden variables and marginalizing them out. We apply the proposed probabilistic embeddings as input to an agglomerative hierarchical clustering (AHC) algorithm to do diarization in the DIHARD19 evaluation set. We compute the full PLDA likelihood by the book for each clustering hypothesis that is considered by AHC. We do joint discriminative training of the PLDA parameters and of the probabilistic x-vector extractor. We demonstrate accuracy gains relative to a baseline AHC algorithm, applied to traditional xvectors (without uncertainty), and which uses averaging of binary log-likelihood-ratios, rather than by-the-book scoring.
@INPROCEEDINGS{FITPUB12288,
author = "Anna Silnova and Langenhoven Nikolaas Johan Brummer and A. Johan Rohdin and Themos Stafylakis and Luk\'{a}\v{s} Burget",
title = "Probabilistic embeddings for speaker diarization",
pages = "24--31",
booktitle = "Proceedings of Odyssey 2020 The Speaker and Language Recognition Workshop",
journal = "Proceedings of Odyssey: The Speaker and Language Recognition Workshop",
volume = 2020,
number = 11,
year = 2020,
location = "Tokyo, JP",
publisher = "International Speech Communication Association",
ISSN = "2312-2846",
doi = "10.21437/Odyssey.2020-4",
language = "english",
url = "https://www.fit.vut.cz/research/publication/12288"
}