Challenging Margin-based Speaker Embedding Extractors By Using The Variational Information Bottleneck

Abstract

Speaker embedding extractors are typically trained using a classification loss over the training speakers. During the last few years, the standard softmax/cross-entropy loss has been replaced by the margin-based losses, yielding significant improvements in speaker recognition accuracy. Motivated by the fact that the margin merely reduces the logit of the target speaker during training, we consider a probabilistic framework that has a similar effect. The variational information bottleneck provides a principled mechanism for making deterministic nodes stochastic, resulting in an implicit reduction of the posterior of the target speaker. We experiment with a wide range of speaker recognition benchmarks and scoring methods and report competitive results to those obtained with the state-of-the-art Additive Angular Margin loss.

Stats

• citations0
• S2 citations—
• github stars0
• HF likes0
• heat score0.00
• arxiv keystafylakis2024challenging

Related papers

• Margin Matters: Towards More Discriminative Deep Neural Network Embeddings For Speaker Recognition (2019)15.25
• Large Margin Softmax Loss For Speaker Verification (2019)14.66
• Angular Softmax Loss For End-to-end Speaker Verification (2018)11.19
• A Study On Angular Based Embedding Learning For Text-independent Speaker Verification (2019)2.26
• Rethinking Session Variability: Leveraging Session Embeddings For Session Robustness In Speaker Verification (2023)5.24
• Within-sample Variability-invariant Loss For Robust Speaker Recognition Under Noisy Environments (2020)11.85
• Improved Large-margin Softmax Loss For Speaker Diarisation (2019)6.34
• Unified Hypersphere Embedding For Speaker Recognition (2018)0.00