Publication Details
Fine-Tuning Self-Supervised Models for Language Identification Using Orthonormal Constraint
CAROFILIS, A.
VANDERREYDT, G.
KHALIL, D.
Madikeri Srikanth
Motlíček Petr, doc. Ing., Ph.D. (DCGM)
SCHUEPBACH, C.
Language Identification, Transformers, Wav2Vec2, fine-tuning, low-resource,
out-of-domain,
Self-supervised models trained with high linguistic diversity, such as the XLS-R
model, can be effectively fine-tuned for the language recognition task.
Typically, a back-end classifier followed by statistics pooling layer are added
during train- ing. Commonly used back-end classifiers require a large num- ber of
parameters to be trained, which is not ideal in limited data conditions. In this
work, we explore smaller parame- ter back-ends using factorized Time Delay Neural
Network (TDNN-F). The TDNN-F architecture is also integrated into Emphasized
Channel Attention, Propagation and Aggregation- TDNN (ECAPA-TDNN) models, termed
ECAPA-TDNN-F, reducing the number of parameters by 30 to 50% absolute, with
competitive accuracies and no change in minimum cost. The results show that the
ECAPA-TDNN-F can be extended to tasks where ECAPA-TDNN is suitable. We also test
the effectiveness of a linear classifier and a variant, the Orthonor- mal linear
classifier, previously used in x-vector type systems. The models are trained with
NIST LRE17 data and evalu- ated on NIST LRE17, LRE22 and the ATCO2 LID datasets.
Both linear classifiers outperform conventional back-ends with improvements in
accuracy between 0.9% and 9.1%
@inproceedings{BUT193354,
author="PRASAD, A. and CAROFILIS, A. and VANDERREYDT, G. and KHALIL, D. and MADIKERI, S. and MOTLÍČEK, P. and SCHUEPBACH, C.",
title="Fine-Tuning Self-Supervised Models for Language Identification Using Orthonormal Constraint",
booktitle="ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings",
year="2024",
pages="11921--11925",
publisher="IEEE Signal Processing Society",
address="Seoul",
doi="10.1109/ICASSP48485.2024.10446751",
isbn="979-8-3503-4485-1",
url="https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10446751"
}