Context Detection and Office/Urban Heartbeat
Syed Ali Hasnain, Roozbeh Jafari

Citation
Syed Ali Hasnain, Roozbeh Jafari. "Context Detection and Office/Urban Heartbeat". Talk or presentation, 14, October, 2015; Poster presented at the 2015 TerraSwarm Annual Meeting.

Abstract
The TerraSwarm envisions trillions of heterogeneous devices connected together. Our aim is to use these heterogeneous sensors and by employing data fusion techniques identify context and determine user's intent. We aim at incorporating adaptive and opportunistic learning when new sensors are introduced, and perform context detection when one or more sensing modalities are missing. Our methods are being tested on an experimental setup that leverages longitudinal data collection in the lab, using a variety of sensors (e.g. accelerometer, gyroscope, RSSI, pressure sensor etc.). The sensors can either be wearable or environmental. We use the data streams from these sensors to find useful information like physical coupling between sensors and possibly identifying activities and context. We have already demonstrated detecting physical interactions using motion sensors leveraging weak signal processing operators to identify similarities and fusing them to create strong classifiers. We also use graph theory to model the coupling between sensors and form interval graphs to exhibit the periods, start time and end time of the couplings. The periodicity in these graphs can provide a 'heartbeat' for the user. We transform interval graphs into strings, and algorithms are being developed to determine Eigen words in polynomial time and detect periodicity (heartbeat) using string matching techniques.

Electronic downloads

Internal. This publication has been marked by the author for TerraSwarm-only distribution, so electronic downloads are not available without logging in.

Citation formats

• HTML

  Syed Ali Hasnain, Roozbeh Jafari. <a
  href="http://www.terraswarm.org/pubs/686.html"
  ><i>Context Detection and Office/Urban
  Heartbeat</i></a>, Talk or presentation,  14,
  October, 2015; Poster presented at the <a
  href="http://www.terraswarm.org/conferences/15/annual"
  
  >2015 TerraSwarm Annual Meeting</a>.

• Plain text

  Syed Ali Hasnain, Roozbeh Jafari. "Context Detection
  and Office/Urban Heartbeat". Talk or presentation,  14,
  October, 2015; Poster presented at the <a
  href="http://www.terraswarm.org/conferences/15/annual"
  
  >2015 TerraSwarm Annual Meeting</a>.

• BibTeX

  @presentation{HasnainJafari15_ContextDetectionOfficeUrbanHeartbeat,
      author = {Syed Ali Hasnain and Roozbeh Jafari},
      title = {Context Detection and Office/Urban Heartbeat},
      day = {14},
      month = {October},
      year = {2015},
      note = {Poster presented at the <a
                href="http://www.terraswarm.org/conferences/15/annual"
                
  >2015 TerraSwarm Annual Meeting</a>},
      abstract = {The TerraSwarm envisions trillions of
                heterogeneous devices connected together. Our aim
                is to use these heterogeneous sensors and by
                employing data fusion techniques identify context
                and determine user's intent. We aim at
                incorporating adaptive and opportunistic learning
                when new sensors are introduced, and perform
                context detection when one or more sensing
                modalities are missing. Our methods are being
                tested on an experimental setup that leverages
                longitudinal data collection in the lab, using a
                variety of sensors (e.g. accelerometer, gyroscope,
                RSSI, pressure sensor etc.). The sensors can
                either be wearable or environmental. We use the
                data streams from these sensors to find useful
                information like physical coupling between sensors
                and possibly identifying activities and context.
                We have already demonstrated detecting physical
                interactions using motion sensors leveraging weak
                signal processing operators to identify
                similarities and fusing them to create strong
                classifiers. We also use graph theory to model the
                coupling between sensors and form interval graphs
                to exhibit the periods, start time and end time of
                the couplings. The periodicity in these graphs can
                provide a 'heartbeat' for the user. We transform
                interval graphs into strings, and algorithms are
                being developed to determine Eigen words in
                polynomial time and detect periodicity (heartbeat)
                using string matching techniques. },
      URL = {http://terraswarm.org/pubs/686.html}
  }
  

Posted by Syed Ali Hasnain on 19 Oct 2015.