Unobtrusive camera-based monitoring of behavioural disturbances in patients with dementia living in long term care facilities
Ryan East, Sophia Mbabaali, Camille Sivelle, Noemie Laguelle, Mohamed-Amine Choukou · 2020
The purpose of this project is to develop a platform for behavioural disturbances recognition in patients with dementia (PwD). Using a combination of pose recognition software and object segmentation, the system will identify hazardous behaviours and potentially dangerous objects. The system was able to recognize a number of human actions including: falling, fighting, throwing, and tantrums. Successfully identifiable objects are far more extensive, but notably include dangerous objects such as knives and scissors.