ePrints@IIScePrints@IISc Home | About | Browse | Latest Additions | Advanced Search | Contact | Help

Self-Supervised Metric Learning with Graph Clustering for Speaker Diarization

Singh, P and Ganapathy, S (2021) Self-Supervised Metric Learning with Graph Clustering for Speaker Diarization. In: 2021 IEEE Automatic Speech Recognition and Understanding Workshop, ASRU 2021, 3 - 17 December 2021, Cartagena, pp. 90-97.

[img]
Preview
PDF
ASRU_2021.pdf - Published Version

Download (1MB) | Preview
Official URL: https://doi.org/10.1109/ASRU51503.2021.9688271

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 self-supervised learning. The metric learning network implements a neural model of the probabilistic linear discriminant analysis (PLDA). The self-supervision 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 self-supervision based metric learning along with the graph-based clustering algorithm, we achieve significant relative improvements of 60 and 7 over the x-vector PLDA agglomerative hierarchical clustering (AHC) approach on AMI and the DIHARD datasets respectively in terms of diarization error rates (DER).

Item Type: Conference Paper
Publication: 2021 IEEE Automatic Speech Recognition and Understanding Workshop, ASRU 2021 - Proceedings
Publisher: Institute of Electrical and Electronics Engineers Inc.
Additional Information: The copyright for this article belongs to the Authors.
Keywords: Clustering algorithms; Computer vision; Discriminant analysis; Graphic methods; Iterative methods; Supervised learning, Clusterings; Graph-based clustering; Metric learning; Neural probabilistic linear discriminant analyse; Path integral; Path integral clustering; Probabilistic linear discriminant analysis; Self-supervised learning; Speaker diarization; X-vector, Embeddings
Department/Centre: Division of Electrical Sciences > Electrical Engineering
Date Deposited: 23 May 2023 04:08
Last Modified: 23 May 2023 04:08
URI: https://eprints.iisc.ac.in/id/eprint/81720

Actions (login required)

View Item View Item