Closed-loop object recognition using reinforcement learning

摘要: Current computer vision systems whose basic methodology is open-loop or filter type typically use image segmentation followed by object recognition algorithms. These are not robust for most real-world applications. In contrast, the system presented here achieves performance using reinforcement learning to induce a mapping from input images corresponding parameters. This accomplished confidence level of model matching as signal team automata search parameters during training. The algorithm part evaluation function gives rise significant improvement automatic generation strategies. verified through experiments on sequences indoor and outdoor color with varying external conditions.