First-Person Camera System to Evaluate Tender Dementia-Care Skill

In this paper, we describe a wearable first-person video (FPV) analysis system for evaluating the skill levels of tender dementia-care technique. Using this system, caregivers can evaluate and elevate their care levels by themselves using the systems' feedbacks. From the FPVs of care sessions taken by wearable cameras worn by caregivers, we obtained the 3D facial distance, pose and eye-contact states between caregivers and receivers by using facial landmark detection and deep neural network (DNN)-based eye contact detection. We applied statistical analysis to these features and developed algorithms that provide scores for tender-care skill. To find and confirm our idea, we conducted chronological study to observe the progression of tender care-skill learning using care learners. First, we took FPVs while care training scenes involving novice caregivers, tender-care experts and middle-level students, and found major behavioural differences among them. Second, we performed the same experiments for the participants before and after training sessions of the care. As the result, we found the same behavioural difference between 1) novices and experts and 2) novices before and after taking training sessions. These results indicate that our FPV-based behavior analysis can evaluate the skill progression of the tender dementia-care.