OPPORTUNITY

People

Alberto Calatroni

Kilian Förster

This project is supervised by Dr. Daniel Roggen and Prof. Gerhard Tröster

Activity and Context Recognition in Opportunistic Sensor Configurations

The objective of OPPORTUNITY is to develop mobile systems to recognize human activity and user context with dynamically varying sensor setups, using goal oriented, cooperative sensing. We refer to such systems as opportunistic, since they take advantage of sensing modalities that just happen to be available, rather than forcing the user to deploy specific, application dependent sensor systems.

The projects aims at generic principles, algorithms and system architectures to reliably recognize complex activities and contexts despite the absence of static assumptions about sensor availability and characteristics in opportunistic systems.

We go about this by a combination of efficient large scale sensing; opportunistic context/activity recognition chain instanciation; dynamic adaptation and autonomous evolution; and validation scenarios.

Our Objective: Dynamic Adaptation and Autonomous Evolution in Opportunistic Sensor Setups

A robust activity and context-recognition system must be capable of operating over a long period of time, exploiting new sources of information as they become available and evolving in an autonomous manner, coping with user variability and changes in the number and type of available sensors.

In particular, wearable and ambient nodes should be trained continuously as new context instances naturally arise and labeling should be carried out ideally with no user intervention. Thereby, the activity recognition system has an intrinsic evolving dynamics that reflects the changing availability of sensors, placement, modality as well as changing user

Our key research directions include:

• Adaptive machine learning algorithms for robust activity recognition in presence of sensor and user variability
• Methods to enable lifelong learning in sensor networks by including online incremental training of classifiers in distributed ContextCells
  
• Technological foundations of ContextCells to enable efficient execution of opportunistic activity recognition methods
  

Publications

2011

• Do I look like my neighbors? A Method to Detect Sensor Data Similarity for open-ended Activity Recognition Systems
  Sarah Neff, Claudia Villalonga, Daniel Roggen and Gerhard Tröster
  in: Eighth International Conference on Networked Sensing Systems (INSS'11), 2011
  [ BiBTeX RIS ]
• Shaping Sensor Node Ensembles according to their Recognition Performance within a Planning-based Context Framework
  Claudia Villalonga, Daniel Roggen and Gerhard Tröster
  in: Eighth International Conference on Networked Sensing Systems (INSS'11), 2011
  [ BiBTeX RIS ]

2010

• Identifying Important Action Primitives for High Level Activity Recognition
  Atif Manzoor, Claudia Villalonga, Alberto Calatroni, Hong-Linh Truong, Daniel Roggen, Schahram Dustdar and Gerhard Tröster
  in: 5th European conference on Smart sensing and context (EuroSSC 2010), Springer-Verlag, 2010
  [ BiBTeX RIS ]
• Recording a complex, multi modal activity data set for context recogntion
  Paul Lukowicz, G. Pirkl, David Bannach, F. Wagner, Alberto Calatroni, Kilian Förster, Thomas Holleczek, M. Rossi, Daniel Roggen, Gerhard Tröster, J. Doppler, C. Holzmann, A. Riener, A. Ferscha and Ricardo Chavarriaga
  in: Workshop on {C}ontext-{S}ystems {D}esign, {E}valuation and {O}ptimisation at {ARCS}, 2010, Hannover, Germany, "", 2010
  [ BiBTeX RIS DOI URL ]

2009

• Bringing Quality of Context into Wearable Human Activity Recognition Systems
  Claudia Villalonga, Daniel Roggen, Clemens Lombriser, Piero Zappi and Gerhard Tröster
  in: First International Workshop on Quality of Context (QuaCon), 2009
  [ BiBTeX RIS ]
• Context Cells: Towards Lifelong Learning in Activity Recognition Systems
  Alberto Calatroni, Claudia Villalonga, Daniel Roggen and Gerhard Tröster
  in: Proceedings of the 4th European Conference on Smart Sensing and Context (EuroSSC), Springer, 2009
  [ BiBTeX RIS PDF ]
• OPPORTUNITY: Activity and context awareness in opportunistic open-ended sensor environments
  Daniel Roggen, Kilian Förster, Alberto Calatroni, Andreas Bulling, Thomas Holleczek, Gerhard Tröster, Paul Lukowicz, Gerald Pirkl, David Bannach, Alois Ferscha, Andreas Riener, Clemens Holzmann, R. Chavarriaga and J. del R. Millán
  in: Proc. of the 1st European Future and Emerging Technologies Conference (FET 2009), Prague, Czech Republic, European commission, 2009
  [ BiBTeX RIS ]
• OPPORTUNITY: Towards opportunistic activity and context recognition systems
  Daniel Roggen, Kilian Förster, Alberto Calatroni, Thomas Holleczek, Y. Fang, Gerhard Tröster, Paul Lukowicz, Gerald Pirkl, David Bannach, Kai Kunze, Alois Ferscha, Clemens Holzmann, Andreas Riener, R. Chavarriaga and J. del R. Millán
  in: Proc. 3rd IEEE WoWMoM Workshop on Autononomic and Opportunistic Communications, 2009
  [ BiBTeX RIS ]

Other contributors

• Dr. Fang Yu
• Piero Zappi
• Claudia Villalonga

Links

OPPORTUNITY Official Website