作者: H.D. Wactlar , A. Bharucha , A.G. Hauptmann , Jiang Gao
关键词: Hidden Markov model 、 Maximum-entropy Markov model 、 Artificial intelligence 、 Nursing homes 、 Forward algorithm 、 Machine learning 、 Subspace topology 、 Pattern recognition 、 Computer science 、 Motion (physics) 、 Markov model
摘要: We describe an algorithm for dining activity analysis in a nursing home. Based on several features, including motion vectors and distance between moving regions the subspace of individual person, hidden Markov model is proposed to characterize different stages activities with certain temporal order. Using HMM model, we are able identify start (and ending) events high accuracy low false positive rate. This approach could be successful assisting caregivers assessments resident's levels over time.