A study of the robustness of raw waveform based speaker embeddings under mismatched conditions

TL;DR

This paper investigates the robustness of raw waveform-based speaker embeddings under mismatched conditions, revealing performance issues and proposing strategies like analytic filters and variational dropout to enhance cross-dataset speaker verification.

Contribution

It provides a comprehensive cross-dataset analysis of raw-waveform speaker embeddings and introduces two novel techniques to improve their robustness in mismatched scenarios.

Findings

Raw-waveform systems degrade more under mismatched conditions than spectral systems.

Using analytic filters improves shift-invariance and robustness.

Variational dropout prevents overfitting of irrelevant features.

Abstract

In this paper, we conduct a cross-dataset study on parametric and non-parametric raw-waveform based speaker embeddings through speaker verification experiments. In general, we observe a more significant performance degradation of these raw-waveform systems compared to spectral based systems. We then propose two strategies to improve the performance of raw-waveform based systems on cross-dataset tests. The first strategy is to change the real-valued filters into analytic filters to ensure shift-invariance. The second strategy is to apply variational dropout to non-parametric filters to prevent them from overfitting irrelevant nuance features.