Discriminative Training of VBx Diarization

článek ve sborníku ve WoS nebo Scopus

angličtina

Bayesian HMM clustering of x-vector sequences (VBx) has be- come a widely adopted diarization baseline model in publications and challenges. It uses an HMM to model speaker turns, a gen- eratively trained probabilistic linear discriminant analysis (PLDA) for speaker distribution modeling, and Bayesian inference to esti- mate the assignment of x-vectors to speakers. This paper presents a new framework for updating the VBx parameters using discrim- inative training, which directly optimizes a predefined loss. We also propose a new loss that better correlates with the diarization error rate compared to binary cross-entropy - the default choice for diarization end-to-end systems. Proof-of-concept results across three datasets (AMI, CALLHOME, and DIHARD II) demonstrate the method's capability of automatically finding hyperparameters, achieving comparable performance to those found by extensive grid search, which typically requires additional hyperparameter behavior knowledge. Moreover, we show that discriminative fine-tuning of PLDA can further improve the model's performance. We release the source code with this publication.

speaker diarization, VBx, clustering, variational Bayes, discriminative training

KLEMENT, D.; DIEZ SÁNCHEZ, M.; LANDINI, F.; BURGET, L.; SILNOVA, A.; DELCROIX, M.; TAWARA, N.

14. 4. 2024

IEEE Signal Processing Society

Seoul

979-8-3503-4485-1

ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings

11871

11875

5

https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10446119

klement_icassp2024_Discriminative_Training_of_VBx_Diarization.pdf