Sounds Asleep - Towards better sleep quality with biometric noise metering and mediation

点击翻译按钮获取中文摘要

Noise can have a significant impact on health and wellbeing, including cardiovascular health, sleep disruption, working cognition, and cumulative hearing damage. Many people in the UK describe their environment as excessively noisy, a significant proportion of the UK population is dissatisfied with their auditory environment (~30%) and sleep disruption in the home environment has been shown to affect both mental and physical health. Noise measures are not readily interpreted by non-specialist audiences, and therefore tend not to encourage proactive adoption by the public. New metrics may help to empower people dealing with noise and interrupted sleep. Advances in affordable, comfortable biophysiological sensors suggest a unique opportunity to explore the relationship between existing acoustic measurement, self-reported responses such as annoyance, and biophysiological measures, creating new, consolidated metrics for noise measurement.This project will also represent the first step towards the possibility of using biofeedback to facilitate noise masking using automated soundscape generation. Individually-adaptive soundscape masking may be effective in mediating the impact of noise - the user would optimally remain in a quality state of sleep while the system responds autonomously. Machine learning can provide a way to use complex combinations of biofeedback as useful control signals for soundscape generation. Such techniques require a large amount of data and mass market wearables are now reaching a level of maturity which can facilitate such experiments.