This paper presents a trace-driven simulation study of two classes of retransmission timeout (RTO) estimators in the context of realtime streaming over the Internet. We explore th...
The context-centered approach to object detection and recognition is based on the intuition that the contextual information of real-world scenes provides relevant information for ...
Abstract. We consider the problem of detecting a large number of different classes of objects in cluttered scenes. We present a learning procedure, based on boosted decision stumps...
Antonio B. Torralba, Kevin P. Murphy, William T. F...
Abstract. Subsystem based generalizations of rough set approximations are investigated. Instead of using an equivalence relation, an arbitrary binary relation is used to construct ...
In this work we propose a general approach for representing uncertainty measures in the framework of t-norm based logics. This approach is extended also to classes of measures lik...