Query Expansion System for the VoxCeleb Speaker Recognition Challenge 2020
arXiv:2011.02882v1
AI Analysis
This work addresses speaker recognition for audio data, but it appears incremental as it builds on existing methods like query expansion and score fusion.
The authors tackled speaker verification in the VoxCeleb challenge by applying query expansion and combining scores from Kaldi x-vectors and ResNet, achieving significant progress over the baseline.
In this report, we describe our submission to the VoxCeleb Speaker Recognition Challenge (VoxSRC) 2020. Two approaches are adopted. One is to apply query expansion on speaker verification, which shows significant progress compared to baseline in the study. Another is to use Kaldi extract x-vector and to combine its Probabilistic Linear Discriminant Analysis (PLDA) score with ResNet score.