Automated human activity recognition in smart room

Abstract

Traditional human activity recognition methods are the statistical pattern recognition techniques, e.g. Nearest Neighbor Rule (NNR), and the state-space methods, e.g. the Hidden Markov Model (HMM). We propose three novel connectionist-based approaches a?? the use of the Elman Network (EN), and two hybrids of Neural Network (NN) and HMM, i.e. HMM-NN and NN-HMM, to recognize ten distinct human activities in a smart room environment. A three-level framework has also been suggested, which first detects and verifies the presence of a person, then tracks the subjecta??s head movement over consecutive frames to extract the difference in coordinates as the feature vector that is invariant to the persona??s sex, race and physique, and finally classifies the activities performed using the three proposed classifiers. Recognition and time-complexity comparisons show that all the three proposed classifiers perform more robustly than the conventional methods and demonstrate their greater potential in realizing recognition of continuous and complex activity in the increasingly popular human-activity-based applications.