Powerset multi-class cross entropy loss for neural speaker diarization.

Alexis Plaquet Hervé Bredin

Published in: INTERSPEECH (2023)

Keyphrases

multi class
cross entropy
speaker diarization
speech recognition
log likelihood
neural network
maximum likelihood
network architecture
language modeling
support vector machine
broadcast news
pairwise
bayesian information criterion
model selection
feature selection
language model
evaluation metrics
machine learning