On training targets for noise-robust voice activity detection

TL;DR

This paper introduces a computationally efficient real-time voice activity detection network that outperforms existing methods by using a segmental voice-to-noise ratio as a training target, enhancing noise robustness.

Contribution

It proposes a new real-time VAD model and demonstrates that segmental VNR as a training target improves noise robustness over traditional methods.

Findings

Segmental VNR outperforms clean speech level as a training target.

Multi-target training further enhances VAD performance.

The proposed VAD achieves state-of-the-art results on public datasets.

Abstract

The task of voice activity detection (VAD) is an often required module in various speech processing, analysis and classification tasks. While state-of-the-art neural network based VADs can achieve great results, they often exceed computational budgets and real-time operating requirements. In this work, we propose a computationally efficient real-time VAD network that achieves state-of-the-art results on several public real recording datasets. We investigate different training targets for the VAD and show that using the segmental voice-to-noise ratio (VNR) is a better and more noise-robust training target than the clean speech level based VAD. We also show that multi-target training improves the performance further.