Vae-based Regularization For Deep Speaker Embedding

Abstract

Deep speaker embedding has achieved state-of-the-art performance in speaker recognition. A potential problem of these embedded vectors (called `x-vectors') are not Gaussian, causing performance degradation with the famous PLDA back-end scoring. In this paper, we propose a regularization approach based on Variational Auto-Encoder (VAE). This model transforms x-vectors to a latent space where mapped latent codes are more Gaussian, hence more suitable for PLDA scoring.

Stats

• citations11
• S2 citations—
• github stars0
• HF likes0
• heat score8.09
• arxiv keyzhang2019vae

Related papers

• Vae-based Domain Adaptation For Speaker Verification (2019)7.50
• Investigation Of Using VAE For I-vector Speaker Verification (2017)0.00
• Gaussian Speaker Embedding Learning For Text-independent Speaker Verification (2020)0.00
• Fast Variational Bayes For Heavy-tailed PLDA Applied To I-vectors And X-vectors (2018)8.35
• Deep Speaker Embedding Learning With Multi-level Pooling For Text-independent Speaker Verification (2019)0.00
• Scoring Of Large-margin Embeddings For Speaker Verification: Cosine Or PLDA? (2022)9.76
• Gaussian-constrained Training For Speaker Verification (2018)8.35
• End-to-end DNN Based Speaker Recognition Inspired By I-vector And PLDA (2017)10.35