Skip to Main content Skip to Navigation
Journal articles

Centroid-aware local discriminative metric learning in speaker verification.

Abstract : We propose a new mechanism to pave the way for efficient learning against class-imbalance and improve representation of identity vector (i-vector) in automatic speaker verification (ASV). The insight is to effectively exploit the inherent structure within ASV corpus — centroid priori. In particular: (1) to ensure learning efficiency against class-imbalance, the centroid-aware balanced boosting sampling is proposed to collect balanced mini-batch; (2) to strengthen local discriminative modeling on the mini-batches, neighborhood component analysis (NCA) and magnet loss (MNL) are adopted in ASV-specific modifications. The integration creates adaptive NCA (AdaNCA) and linear MNL (LMNL). Numerical results show that LMNL is a competitive candidate for low-dimensional projection on i-vector (EER = 3.84% on SRE2008, EER = 1.81% on SRE2010), enjoying competitive edge over linear discriminant analysis (LDA). AdaNCA (EER = 4.03% on SRE2008, EER = 2.05% on SRE2010) also performs well. Furthermore, to facilitate the future study on boosting sampling, connections between boosting sampling, hinge loss and data augmentation have been established, which help understand the behavior of boosting sampling further.
Complete list of metadata

Cited literature [40 references]  Display  Hide  Download
Contributor : Joseph Razik Connect in order to contact the contributor
Submitted on : Thursday, May 17, 2018 - 2:39:05 PM
Last modification on : Thursday, July 14, 2022 - 4:07:12 AM
Long-term archiving on: : Tuesday, September 25, 2018 - 9:05:55 PM


Files produced by the author(s)




Kekai Sheng, Weiming Dong, Wei Li, Joseph Razik, Feiyue Huang, et al.. Centroid-aware local discriminative metric learning in speaker verification.. Pattern Recognition, Elsevier, 2017, 72 (c), pp.176-185. ⟨10.1016/j.patcog.2017.07.007⟩. ⟨hal-01769892⟩



Record views


Files downloads