Irina Mocanu

Human Activity Recognition in Smart Environments

This paper presents a method for image based human activity recognition, in a smart environment. We use background subtraction and skeletisation as image processing techniques, combined with Artificial Neural Networks for human posture classification and Hidden Markov Models for activity interpretation. By this approach we successfully recognized basic human actions such as walking, rotating, sitting and bending up/down, lying and falling. The method can be applied in smart houses, for elderly people who live alone.

A Multi-agent System for Human Activity Recognition in Smart Environments

Activity recognition is an important component for the ambient assisted living systems, which perform home monitoring and assistance of elderly people or patients with risk factors. The paper presents a prototype system for activity recognition based on a multi-agent architecture. In the system, the context of the person is first detected using a domain ontology. Next, the human position is obtained and together with the context forms a sub-activity. The sequence of successive sub-activities is then assembled in a human activity, which is recognized using a stochastic grammar.

Human Activity Recognition Based on Multiple Kinects

Activity recognition is an important component for the ambient assisted living systems which perform home monitoring and assistance for elderly people or patients with risk factors. This paper presents a prototype system for activity recognition using information provided by four Kinects. First the posture of the supervised person is detected using a set of rules created with ID3 algorithm applied to a skeleton obtained by merging the skeletons provided by multiple Kinects. At the same time, the interaction of the user with the objects from the house is determined. After that, daily activities are identified using Hidden Markov Models in which the detected postures and the object interactions are observable states. The benefit of merging the information received from multiple Kinects together with the detection of the interaction between the user and relevant objects from the room is the increase in accuracy for the recognized activities.

A Multi-Agent System for Service Acquiring in Smart Environments

This paper proposes a multi-agent system architecture for a service acquiring system. The proposed system is integrated in an ambient intelligent system, AmIHomCare, a smart house who supervises elderly people in their homes and also helps people during their daily activities. The service acquiring system is specified and validated using a formal specification method - Event B.

Genetic Algorithms Viewed as Anticipatory Systems

This paper proposes a new version of genetic algorithms—the anticipatory genetic algorithm AGA. The performance evaluation included in the paper shows that AGA is superior to traditional genetic algorithm from both speed and accuracy points of view. The paper also presents how this algorithm can be applied to solve a complex problem: image annotation, intended to be used in content based image retrieval systems.

Data Fusion in a Multi Agent System for Person Detection and Tracking in an Intelligent Room

The main components of a supervising system is detecting and tracking of the supervised person in an intelligent room. This paper presents architecture for a non-intrusive multi-agent system for person detection and tracking. The main objective of this system is to offer continuity over the user’s movement, as it can be controlled in such a way so as to keep the user inside the frame for most of the time. The proposed architecture will integrate different types of sensors: multiple Kinect sensors and a PTZ camera, in order to minimize the drawbacks of using only one type of sensor. For example field of view provided by Kinect sensor is not wide enough to cover the entire room. Also the PTZ camera is not able to detect and track a person in case of different special situations, such as the person is sitting or it is under the camera. Furthermore Kinect sensors will help the PTZ camera to control the camera’s orientation. Person detection and tracking is performed using computer vision techniques applied to RGB images. The system is designed over the existing platform AmI-Platform and is partially evaluated in the AmI-Lab laboratory from the University Politehnica of Bucharest (UPB).

A multi-agent supervising system for smart environments

This paper presents a multi-agent architecture for a supervising system which is part of a new intelligent ambient system, called AmIHomCare. The supervising system detects the context of the person (using a domain ontology): the person location in the room together with the surrounding objects and its body posture. A sequence of these contexts is assembled in a daily activity, based on a set of rules, described by a stochastic context free grammar with attributes. Also the system predicts the next context and the next activity in order to detect emergencies in the smart environment.

A Kinect based adaptive exergame

This paper describes an application that aims to stimulate physical activity adapted to elderly people. The application is implemented as a game with two avatars: user and trainer. The user avatar must reproduce the movements of the trainers avatar. The similarity between two movements is computed using Dynamic Time Warping. Also the speed of the trainer is adapted to the users movements using cross-correlation. Thus the game can stimulate for a long time physically activity. The preliminary evaluations with ten people have shown that the system can be an effective tool that engages users into physical activity.

A Novel Integrated Architecture for Ambient Assisted Living Systems

The increase in life expectancy and the slumping birth rates across the world result in lengthening the average age of the society. Therefore, we are in need of techniques that will assist the elderly in their daily life, while preventing their social isolation. The recent developments in Ambient Intelligence and Information and Communication Technologies have facilitated a technological revolution in the field of Ambient Assisted Living. At present, there are many technologies on the market that support the independent life of older adults, requiring less assistance from family and caregivers, yet most of them offer isolated services, such as health monitoring, reminders etc, moreover none of current solutions incorporates the integration of various functionalities and user preferences or are formally analyzed for their functionality and quality-of-service attributes, a much needed endeavor in order to ensure safe mitigations of potential critical scenarios. In this paper, we propose a novel architectural solution that integrates necessary functions of an AAL system seamlessly, based on user preferences. To enable the first level of the architectures analysis, we model our system in Architecture Analysis and Design Language, and carry out its simulation for analyzing the end-to-end data-flow latency, resource budgets and system safety.

Agent-Based System for Affective Intelligent Environment

Within the Ambient Intelligence research field there is little attention payed to the affective side of the users. However, we believe that it should not be ignored, because emotions have an important role in all human cognitive processes and in their behavior. A smart environment should be able to detect the emotional state of the persons living there and to adjust its answers according to the user affective needs.