Abstract | ||
---|---|---|
The last decade has seen a considerable increase in the number of sensors we interact with on a daily basis. However, it is not always possible for a single sensing system to capture the complete story. While statically mounted infrastructure sensors typically capture the what, where, how much etc aspects of a detected event, e.g. (what appliance was used, how much energy did it consume), they do not always answer the who question. On the other hand, the advent of wearables has helped answer the what and who aspects - e.g. (who used the appliance). Fusing such sensor streams that observe the same event but different attributes of it, opens up many interesting applications. In this paper, we present a globally optimal data fusion algorithm for such pairs of systems, and show why traditional bipartite algorithms do not work. We evaluate our algorithm against two greedy baselines and show that our algorithm has lesser variance in the presence of time skew, false positives and false negatives. |
Year | DOI | Venue |
---|---|---|
2015 | 10.1145/2800835.2801630 | UbiComp/ISWC Adjunct |
Field | DocType | Citations |
Data mining,Computer science,Wearable computer,Bipartite graph,Baseline (configuration management),Sensor fusion,Data fusion algorithms,Skew,False positives and false negatives,Wireless sensor network,Embedded system | Conference | 3 |
PageRank | References | Authors |
0.38 | 7 | 2 |
Name | Order | Citations | PageRank |
---|---|---|---|
Avinash Kalyanaraman | 1 | 34 | 4.81 |
Kamin Whitehouse | 2 | 1982 | 160.79 |