Control of sensory perception in discrete event systems using stochastic dynamic programming

The problem of controlling sensory perception for use in discrete event feedback control systems is addressed in this paper. The problem consists of the real-time dynamic selection of different discrete event recognition techniques, which produce confidence levels of the recognized events. For a discrete event control system running in normal operation, the confidence levels are typically large and only a few event recognizers are needed. Then, as the event recognition becomes harder, the confidence levels will decrease and additional event recognizers are utilized by the sensory perception controller. The final product is an intelligent architecture with the ability to actively control the use of sensory input and perception to achieve high performance discrete event recognition.