Handling camera movement constraints in reinforcement learning based active object recognition
DAGM'06 Proceedings of the 28th conference on Pattern Recognition
Hi-index | 0.00 |
We discuss a rigorous unifying framework for both planning andreplanning, extending an existing logic-based approach toresource-based planning. The primitive concepts in this ActionResource Framework (ARF) are actions and resources. Actions consumeand produce resources. Plansare structures composed of actions,resource facts and an explicit dependency function specifying theirinterrelationships. In this framework, both planning and replanningare conceived as plan transformation processes accomplished byapplying sequences of operations on plans. For this, we introduceoperators for plan transformation and define the concept of a planlibrary. Using a refinement planning template, we show how someexisting (re)planning methods and heuristics can be described asspecial cases of this framework. The advantage of the framework isthat it offers a unifying view on planning and replanning.