Research Output per year
Phd Researcher Profile
Thesis Title: Video analysis for Detection of Abnormal Behaviour in Smart Environments
Thesis Abstract: We aim to develop a machine learning classification model which uses data captured by video to provide ongoing automatic recognition of agitation in people living with dementia in a way that is sensitive to the needs of the user in order to supplement existing health metrics and enable timely intervention by carers.
Agitation may lead people living with dementia to cause harm to themselves, their carers or others, necessitating the use of physiological or pharmacological restraints. Early detection can ease carer burden and prevent symptoms escalating, further distress, the use of restraints or injury. Automatic detection of agitation can provide a fuller picture of a PwD’s health and behaviour and enable appropriate carer intervention. Automatic detection has the potential to increase the assessment accuracy by supplementing information and removing human subjectivity. Our work is be based upon the Brief Agitation Report Scale, a widely used measure of agitation.
Our research seeks to create a model that uses a patient’s skeletal joint location, which will be obtained using a conventional video camera, to detect agitated behaviour.
Bachelor, Heriot-Watt University
Research output: Chapter in Book/Report/Conference proceeding › Conference contribution