SelfSupervised Metric Learning With Graph Clustering For Speaker Diarization
Abstract
In this paper, we propose a novel algorithm for speaker diarization using metric learning for graph based clustering. The graph clustering algorithms use an adjacency matrix consisting of similarity scores. These scores are computed between speaker embeddings extracted from pairs of audio segments within the given recording. In this paper, we propose an approach that jointly learns the speaker embeddings and the similarity metric using principles of selfsupervised learning. The metric learning network implements a neural model of the probabilistic linear discriminant analysis (PLDA). The selfsupervision is derived from the pseudo labels obtained from a previous iteration of clustering. The entire model of representation learning and metric learning is trained with a binary cross entropy loss. By combining the selfsupervision based metric learning along with the graphbased clustering algorithm, we achieve significant relative improvements of 60% and 7% over the xvector PLDA agglomerative hierarchical clustering (AHC) approach on AMI and the DIHARD datasets respectively in terms of diarization error rates (DER).
 Publication:

arXiv eprints
 Pub Date:
 September 2021
 arXiv:
 arXiv:2109.06824
 Bibcode:
 2021arXiv210906824S
 Keywords:

 Electrical Engineering and Systems Science  Audio and Speech Processing;
 Computer Science  Sound
 EPrint:
 8 pages, Accepted in ASRU 2021