A5089575499- Context-Aware Activity Recognition Systems
- Interactive and Immersive Displays
- Simulation Techniques and Applications
- Gaze Tracking and Assistive Technology
- Evacuation and Crowd Dynamics
- Tactile and Sensory Interactions
- Human-Automation Interaction and Safety
- IoT and Edge/Fog Computing
- Distributed and Parallel Computing Systems
- Business Process Modeling and Analysis
- Human Mobility and Location-Based Analysis
- Modular Robots and Swarm Intelligence
- Petri Nets in System Modeling
- Robotics and Automated Systems
- Personal Information Management and User Behavior
- Augmented Reality Applications
- Energy Efficient Wireless Sensor Networks
- Mobile Crowdsensing and Crowdsourcing
- Data Visualization and Analytics
- Distributed systems and fault tolerance
- Transportation Planning and Optimization
- Healthcare Technology and Patient Monitoring
- Anomaly Detection Techniques and Applications
- Mobile Agent-Based Network Management
- Innovative Human-Technology Interaction
Johannes Kepler University of Linz
2016-2025
Procomcure Biotech (Austria)
2019-2021
Research Studios Austria
2008-2019
A.S. Watson (Netherlands)
2014
Carnegie Mellon University
2014
Microsoft Research (United Kingdom)
2014
University of Duisburg-Essen
2014
Private University College of Education of the Diocese of Linz
2010
Software Competence Center Hagenberg (Austria)
2009
University of Passau
2009
We deployed 72 sensors of 10 modalities in 15 wireless and wired networked sensor systems the environment, objects, on body to create a sensor-rich environment for machine recognition human activities. acquired data from 12 subjects performing morning activities, yielding over 25 hours data. report number activity occurrences observed during post-processing, estimate that 13000 14000 object interactions occurred. describe setup methodology acquisition, synchronization curation. challenges...
The new generation of smartphones has realized the early vision context awareness. next step is facilitating real-world impact more complex recognition, moving toward next-generation opportunistic recognition configurations and large-scale ensembles networked subsystems interacting with communities users.
Achieving true ambient intelligence calls for a new opportunistic activity recognition paradigm in which, instead of deploying information sources specific goal, the methods themselves dynamically adapt to available sensor data. The Web extra at http://youtu.be/dMN3Hoq5QfU is video recording configurations highly rich environment that shows how large number instances occur during early morning activities an apartment.
Learning underlying patterns from sensory data is crucial in the Human Activity Recognition (HAR) task to avoid poor generalization when coping with unseen data. A key solution such an issue representation learning, which becomes essential input signals contain activities similar or generated by different subjects for same activity vary. To address these issues, we seek a increase learning factors of each sensor signal. We develop novel multi-channel asymmetric auto-encoder recreate...
Opportunistic sensing allows to efficiently collect information about the physical world and persons behaving in it. This may mainstream human context activity recognition wearable pervasive computing by removing requirements for a specific deployed infrastructure. In this paper we introduce newly started European research project OPPORTUNITY within which develop mobile opportunistic systems. We outline project's objective, approach follow along sensing, data processing interpretation,...
Vibro-tactile stimulation has been revealed as a potentially effective means to deliver spontaneous notifications like alerts recipients that are focused on other tasks (although only at very low bit rates, and depending the place which tactors placed). This work addresses issue of amount information can be perceived via stimuli coming from wrist worn tactors, given recipient is not expecting or attentive potential occurrence an alert. Assuming apparel watches with embedded represent alert...
Here we present the overall objectives and approach of SAPERE ("Self-aware Pervasive Service Ecosystems") project, focussed on development a highly-innovative nature-inspired framework, suited for decentralized deployment, execution, management, self-aware adaptive pervasive services in future network scenarios.
Driven by technological advancements, gesture interfaces have recently found their way into vehicular prototypes of various kind. Unfortunately, application is less than perfect and detailed information about preferred execution regions, spatial extent, time behavior are not available yet. Providing car (interior) manufacturer with characteristics would allow them to design future in-vehicle concepts in a interfere gestural interaction. To tackle the problem, this research aims as...
In a distributed memory environment the communication overhead of Time Warp as induced by rollback procedure due to “overoptimistic” progression simulation is dominating performance factor. To limit optimism an extent that can be justified from inherent model parallelism, control mechanism proposed, which maintaining history record virtual time differences stamps carried arriving messages, and forecasting timestamps forthcoming probabilistically delays execution scheduled events avoid...
Driving a vehicle is task affected by an increasing number and rising complexity of Driver Assistance Systems (DAS) resulting in raised cognitive load the driver, consequence to distraction from main activity driving. A potential solutions have been proposed so far, however, although these techniques broaden perception horizon (e. g. introduction sense touch as additional information modality or utilization multimodal instead unimodal interfaces), they demand attention driver too. In order...
In alpine skiing, accurate and real-time estimation of body pose inclinations due to turning is critical as it demonstrates the skier’s behavior abilities. Although inertial measurement units (IMUs) ease measuring kinematics in extreme conditions provide such indications skiers’ behavior, they often suffer from sensor placement orientation variability. This study explains impact on captured signals proposes a preprocessing algorithm that can rotate raw various locations orientations similar...
The main performance pitfall of the Time Warp distributed discrete event simulation (DDES) protocol has been widely recognized to be overoptimistic progression execution into simulated future. premature events that eventually have "rolled back" due causality violations induces memory and communication overheads as sources inefficiencies. Optimistic Windows self adaptive mechanisms proposed in literature control optimism order improve or optimize its performance. An mechanism based on...
To increase crowd evacuation efficiency in emergency situations, the authors developed a wearable computer, LifeBelt, to guide individuals from situations escape. The LifeBelt collects position and orientation information about embedded sensors, shares this with background coordination system, receives an individualized optimized escape strategy, guides individual via vibro-tactile stimulation, indicating direction distance recommended exit. Simulation models indicate that number of...
Opportunistic sensing can be used to obtain data from sensors that just happen present in the user’s surroundings. By harnessing these opportunistic sensor configurations infer activity or context, ambient intelligence environments become more robust, have improved user comfort thanks reduced requirements on body-worn deployment and they are not limited a predefined restricted location, defined by specifically deployed for an application. We OPPORTUNITY Framework Data Processing Ecosystem...
Collective Adaptive System (CAS) as a broad term refers to todays observable indications of deployment Information and Communication Technologies (ICT) at very large scale (e.g. 3.5*109 personal computers, 6.2*109 smart phones, 1.1*109 cars on planet). Taking such massive ICT deployment, it is not just considered possible, but already reality that these are programmed operate cooperatively collectives. As for example, smartphone apps have demonstrated globe-spanning cooperative sensing...
Distributed simulations of timed transition Petri nets on real multiprocessor systems can be faster than sequential simulation, even for small simulation models. The authors show how.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">></ETX>