A Comparison of Perceptually Motivated Loss Functions for Binary Mask Estimation in Speech Separation.

Danny Websdale Ben Milner

Published in: INTERSPEECH (2017)

Keyphrases

loss function
pairwise
support vector
logistic regression
squared error
loss minimization
risk minimization
learning to rank
human perception
reproducing kernel hilbert space
convex loss functions
similarity measure
language model
parameter estimation
hinge loss