A vision-based automatic tracking system for ocean animals in the midwater has been demonstrated in Monterey Bay, CA. Currently, the input to this system is a measurement of relati...
Particle filtering algorithms can be used for the monitoring of dynamic systems with continuous state variables and without any constraints on the form of the probability distribu...
Particle filters (PFs) are powerful samplingbased inference/learning algorithms for dynamic Bayesian networks (DBNs). They allow us to treat, in a principled way, any type of prob...
Arnaud Doucet, Nando de Freitas, Kevin P. Murphy, ...
This work extends and improves a recently introduced (Dec. 2007) dynamic Bayesian network (DBN) based audio-visual automatic speech recognition (AVASR) system. That system models ...
Numerous temporal inference tasks such as fault monitoring and anomaly detection exhibit a persistence property: for example, if something breaks, it stays broken until an interve...