logo Idiap Research Institute        
 [BibTeX] [Marc21]
Robust and Discriminative Speaker Embedding via Intra-Class Distance Variance Regularization
Type of publication: Conference paper
Citation: Le_INTERSPEECH2018_2018
Publication status: Published
Booktitle: Proceedings of Interspeech
Year: 2018
Pages: 2257-2261
Location: Hyderabad, INDIA
ISSN: 2308-457X
ISBN: 978-1-5108-7221-9
DOI: 10.21437/Interspeech.2018-1685
Abstract: Learning a good speaker embedding is critical for many speech processing tasks, including recognition, verification, and diarization. To this end, we propose a complementary optimizing goal called intra-class loss to improve deep speaker embed dings learned with triplet loss. This loss function is formulated as a soft constraint on the averaged pair-wise distance between samples from the same class. Its goal is to prevent the scattering of these samples within the embedding space to increase the intra-class compactncss.When intra-class loss is jointly optimized with triplet loss, we can observe 2 major improvements: the deep embedding network can achieve a more robust and discriminative representation and the training process is more stable with a faster convergence rate. We conduct experiments on 2 large public benchmarking datasets for speaker verification, VoxCeleb and VoxForge. The results show that intra-class loss helps accelerating the convergence of deep network training and significantly improves the overall performance of the resulted embeddings.
Keywords: deep neural networks, embedding learning, speaker verification, triplet loss
Projects Idiap
Authors Le, Nam
Odobez, Jean-Marc
Added by: [UNK]
Total mark: 0
  • Le_INTERSPEECH2018_2018.pdf