Recognition of Daily Activity Patterns with Associative Memory and Recall Model

As a smart home, it should be able to be aware of human daily living activities at home. With the advent of IoT, interconnected devices are becoming pervasive and deployed in a smart home. These devices include IP cameras, pressure sensors, air conditioners etc. typically, data from these devices is processed by a central controller to understand situations and take actions. A lot of research effort has been dedicated to creating enhanced algorithm for central controllers to make better decisions. Most of them are typically based on pattern analysis, clustering or statistics techniques. In this study, we propose a cognitive controller which encompasses knowledge acquisition, learning and recalling cognitive functions just as in the human brain by employing an associative memory and recall model (AMR) in conjunction with the KID model. Having a cognitive controller, a smart home system can work efficiently and effectively in recognition of human daily living activity. In this study, the Activities of Daily Living (ADL) dataset is used, the KID and the AMR based smart system is implemented for learning and identifying some cases of human daily living activities.