Acoustic Sounds for Wellbeing: A Novel Dataset and Baseline Results
This work addresses the need for methods to improve wellbeing in noisy urban environments by providing a dataset for sound healing practices, but it is incremental as it focuses on dataset creation and baseline results.
The study introduced the Acoustic Sounds for Wellbeing (ASW) dataset, containing over 88 hours of audio from five classes of acoustic instrumentation, and reported baseline classification results with a maximum unweighted average recall of 57.4% using conventional features and a support vector machine.
The field of sound healing includes ancient practices coming from a broad range of cultures. Across such practices there is a variety of acoustic instrumentation utilised. Practitioners suggest that sound has the ability to target both mental and even physical health issues, e.g., chronic-stress, or joint-pain. Instruments including the Tibetan singing bowl and vocal chanting, are still widely used today. With the noise-floor of modern urban soundscapes continually increasing and known to impact wellbeing, methods to improve this are needed. With that in mind, this study presents the Acoustic Sounds for Wellbeing (ASW) dataset. The ASW dataset is a dataset gathered from YouTube including 88\,+ hrs of audio from 5-classes of acoustic instrumentation (Gongs, Drumming, Singing Bowls, and Chanting). We additionally present initial baseline classification results on the dataset, finding that conventional Mel-Frequency Cepstra coefficient features achieve at best an unweighted average recalled of 57.4 % for a 5-class support vector machine classification paradigm.