Activity Recognition and Abnormality Detection with the Switching Hidden Semi-Markov Model

Socially intelligent surveillance and monitoring: Analysing social dimensions of physical space

“…For example, in the early analysis proposed in [8], the behavior is a hierarchical entity tightly connected with the notion of human motion or action; More specifically, a movement is the most basic brick, requiring no contextual or sequential knowledge to be recognized; the action is a larger scale event formed by ordered movements, which typically includes interaction with the environment and causal relationships. Similar structural definitions can be found in several works [34,2,9,59,35,18]. In [54], human behaviors is accurately described as a set of dynamic models (e.g., Kalman filters) sequenced together by using a Markov chain.…”

“…In almost every surveillance system published in literature, the main goal is that of promptly identifying threatening behaviours in an automatic way. Assuming that the meaning of behavior has been grounded, the underlying, subtle, problem is that the meaning of "threatening" is actually unspecified, and reduced (too) often to that of "abnormal" or "unexpected" [62,18]. This translates in having complex techniques that simply collect a statistics of trajectories, and whenever a different trajectory does hold, it is labelled as abnormal and considered as potentially threatening.…”

“…In [25,35], time duration of actions is explicitly taken into account employing a Markov logic approach. A hierarchical notion of behavior, that enriches the serial action-based definition by adding different abstraction levels, is proposed in [18,49]. In [17], a completely novel concept of behavior is proposed, which consists of an ensemble of spatiotemporal feature points, deeply investigated in [50].…”

“…Some of them avoid to focus on an operational translation of behavior [11,70], often employing the notions of behavior and activity interchangeably. Other approaches often propose ad-hoc definitions, well-suited for the task at hand [8,54,34,2,9,17,59,35,18], generating a bunch of diverse characterizations. For example, in the early analysis proposed in [8], the behavior is a hierarchical entity tightly connected with the notion of human motion or action; More specifically, a movement is the most basic brick, requiring no contextual or sequential knowledge to be recognized; the action is a larger scale event formed by ordered movements, which typically includes interaction with the environment and causal relationships.…”

Activity recognition in smart homes with self verification of assignments

“…A dense sensing platform is used, such as contact switch sensors to monitor the opening and closing of doors, motion sensors to detect the user presence at a particular location, or pressure sensors to indicate the usage of objects, bed or sofa [12,31,32]. Switch sensors deployed in multiple objects in a home such as doors, windows, cupboards and refrigerator, can be used in the NB classifier based recognition approaches [12,30], where NB identifies the activity corresponding to the sensor values with the highest probability.…”