Maziyar Baran Pouyan, Javad Birjandtalab1, Mehrdad Heydarzadeh2, Mehrdad Nourani, Sarah Ostadabbas3
13:25 - 14:15 | Thu 16 Feb | Ballroom D | ThRPF.8
Monitoring sleep postures can provide critical information when analyzing an individual's sleep quality and in-bed behavior. Furthermore, tracking sleep posture over time can play an important role in preventing pressure ulcers (bed-sores) in bed-bound patients who are unable to move and change their position frequently. Pressure sensing mats consist of gridded and flexible force sensors are now commercially available for continuously measuring pressure distribution under body parts in different in-bed postures. In this paper, we report the results of a data collection study conducted in two separate experimental sessions from 13 participants in various sleeping postures using two commercial pressure mats. This resource, released publicly, would benefit future research in the area of sleep behavior/quality and corresponding complications. Moreover, we have employed an algorithm based on deep learning for subject identification in the three common sleeping postures using statistical features extracted from the pressure distribution. Our experiments showed promising results in subject identification and further validated the personal sleeping style of each participant.