Fast Variational Bayes For Heavy-tailed PLDA Applied To I-vectors And X-vectors

Abstract

The standard state-of-the-art backend for text-independent speaker recognizers that use i-vectors or x-vectors, is Gaussian PLDA (G-PLDA), assisted by a Gaussianization step involving length normalization. G-PLDA can be trained with both generative or discriminative methods. It has long been known that heavy-tailed PLDA (HT-PLDA), applied without length normalization, gives similar accuracy, but at considerable extra computational cost. We have recently introduced a fast scoring algorithm for a discriminatively trained HT-PLDA backend. This paper extends that work by introducing a fast, variational Bayes, generative training algorithm. We compare old and new backends, with and without length-normalization, with i-vectors and x-vectors, on SRE'10, SRE'16 and SITW.

Stats

• citations12
• S2 citations—
• github stars0
• HF likes0
• heat score8.35
• arxiv keysilnova2018fast

Related papers

• Gaussian Speaker Embedding Learning For Text-independent Speaker Verification (2020)0.00
• Vae-based Regularization For Deep Speaker Embedding (2019)8.09
• Investigation Of Using VAE For I-vector Speaker Verification (2017)0.00
• Gaussian Meta-embeddings For Efficient Scoring Of A Heavy-tailed PLDA Model (2018)8.09
• Joint Bayesian Gaussian Discriminant Analysis For Speaker Verification (2016)3.58
• Probabilistic Spherical Discriminant Analysis: An Alternative To PLDA For Length-normalized Embeddings (2022)6.77
• Multiobjective Optimization Training Of PLDA For Speaker Verification (2018)2.26
• Language-depedent I-vectors For LRE15 (2017)0.00