Snore-GANs: Improving Automatic Snore Sound Classification with Synthesized Data
This addresses a data scarcity problem for researchers and practitioners in medical audio analysis, but it is incremental as it applies an existing GAN framework to a specific domain.
The paper tackles the lack of supervised training data in automatic snore sound classification by proposing a semi-supervised conditional GAN approach for data augmentation, which outperforms classic augmentation methods and is competitive with recent systems.
One of the frontier issues that severely hamper the development of automatic snore sound classification (ASSC) associates to the lack of sufficient supervised training data. To cope with this problem, we propose a novel data augmentation approach based on semi-supervised conditional Generative Adversarial Networks (scGANs), which aims to automatically learn a mapping strategy from a random noise space to original data distribution. The proposed approach has the capability of well synthesizing 'realistic' high-dimensional data, while requiring no additional annotation process. To handle the mode collapse problem of GANs, we further introduce an ensemble strategy to enhance the diversity of the generated data. The systematic experiments conducted on a widely used Munich-Passau snore sound corpus demonstrate that the scGANs-based systems can remarkably outperform other classic data augmentation systems, and are also competitive to other recently reported systems for ASSC.