Deep Neural Network Voice Activity Detector for Downsampled Audio Data: An Experiment Report
This work addresses the challenge of analyzing team interactions using downsampled audio, but it is incremental with limited practical impact.
The researchers tackled the problem of detecting voice activity in downsampled audio data from sociometric badges, achieving only moderate accuracy in low-noise and variable-noise settings despite good validation performance.
Sociometric badges are an emerging technology for study how teams interact in physical places. Audio data recorded by sociometric badges is often downsampled to not record discussions of the sociometric badges holders. To gain more information about interactions inside teams with sociometric badges a Voice Activity Detector (VAD) is deployed to measure verbal activity of the interaction. Detecting voice activity from downsampled audio data is challenging because down-sampling destroys information from the data. We developed a VAD using deep learning techniques that achieves only moderate accuracy in a low noise meeting setting and in across variable noise levels despite excellent validation performance. Experiences and lessons learned while developing the VAD are discussed.