First-person Vision-based Assessment of Fall Risks in The Wild, Towards Fall Prevention in Older Adults

Falls in older adults is one of the most important public health problems world-wide. In our previous works, we showed that first-personvision (FPV) data acquired by chest- and waist-mounted camerashave the potential to be utilized to (A) develop novel markerlessdeep models to estimate spatiotemporal gait parameters over time(e.g., step width) by localizing feet in 2D coordinate system of RGBframes (using optical flow and RGB streams) and (B) automaticallyidentify environmental hazards (e.g., curbs, stairs, different terrains)that may lead to falling. In this paper, a summary of our recent FPV-based approaches for fall risk assessment in the wild are being discussed. These approaches aimed to eventually inform clinical decisions on the most appropriate prevention interventions to reducefall incidence in older populations.