Exploiting prior knowledge, we use Bayesian estimation to localize a source heard by a fixed sensor network. The method has two main aspects: Firstly, the probability density fun...
This paper proposes a unified framework for spatiotemporal segmentation of video sequences. A Bayesian network is presented to model the interactions among the motion vector field...
Most of the current Intrusion Detection Systems (IDS) examine all data features to detect intrusion or misuse patterns. Some of the features may be redundant or contribute little (...
Srilatha Chebrolu, Ajith Abraham, Johnson P. Thoma...
Abstract We argue that in the decision making process required for selecting assertible vague descriptions of an object, it is practical that communicating agents adopt an epistemi...
When high-dimensional data vectors are visualized on a two- or three-dimensional display, the goal is that two vectors close to each other in the multi-dimensional space should als...
Petri Kontkanen, Jussi Lahtinen, Petri Myllymä...