In this paper, we present our approach and results of the MediaEval 2011 social event detection (SED) task. We solve the event detection problem in three steps. First, we query all event instances that happened given some condition. Then, an event identi cation model is proposed to measure the relationship between events and photos. Finally, visual pruning and owner re ning heuristics are employed to improve the results.
Type:
Conference
City:
Pisa
Date:
2011-09-01
Department:
Data Science
Eurecom Ref:
3519
Copyright:
© EURECOM. Personal use of this material is permitted. The definitive version of this paper was published in and is available at :
See also:
