Fine-tuning Self-Supervised Models For Language Identification Using Orthonormal Constraint
| Type of publication: | Conference paper |
| Citation: | Prasad_ICASSP2024_2024 |
| Publication status: | Published |
| Booktitle: | Proceedings of the 49th IEEE International Conference on Acoustics, Speech, & Signal Processing (ICASSP) |
| Year: | 2024 |
| Pages: | 11921-11925 |
| ISSN: | 2379-190X |
| DOI: | 10.1109/ICASSP48485.2024.10446751 |
| Abstract: | 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 training. Commonly used back-end classifiers require a large number of parameters to be trained, which is not ideal in limited data conditions. In this work, we explore smaller parameter 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 Orthonormal linear classifier, previously used in x-vector type systems. The models are trained with NIST LRE17 data and evaluated 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%. |
| Keywords: | |
| Projects |
Idiap armasuisse |
| Authors | |
| Added by: | [UNK] |
| Total mark: | 0 |
|
Attachments
|
|
|
|
|
Notes
|
|
|
|
|
