Application of Adversarial Domain Adaptation to Voice Activity Detection

Voice Activity Detection (VAD) is becoming an essential front-end component in various speech processing systems. As those systems are commonly deployed in environments with diverse noise types and low signal-to-noise ratios (SNRs), an effective VAD method should perform robust detection of speech region out of noisy background signals. In this paper, we propose applying an adversarial domain adaptation technique to VAD. The proposed method trains DNN models for a VAD task in a supervised manner, simultaneously mitigating the problem of area mismatch between noisy and clean audio stream in a unsupervised manner. The experimental results show that the proposed method improves robust detection performance in noisy environments compared to other DNN-based model learned with hand-crafted acoustic feature.