选择性注意计算模型与算法发展综述

A Survey on Selective Attention Computing Models and Algorithms

人类在利用视觉进行环境感知时，能够从纷繁复杂的场景中快速、准确地提取有用信息，这是因为人类自然视觉具有选择性注意这一重要特征。为了模拟并实现自然视觉的这一特征，在计算机视觉领域，学者们研究并提出了诸多选择性注意的计算模型。在此对已有的选择性注意计算模型与算法进行了综述，介绍了国内外研究现状，提出了现有研究存在的不足。建议进一步研究自下而上基于图像特征的信息以及自上而下基于任务、知识的信息对视觉注意力的驱动机制，研究选择性注意非均匀采样的表现形式及其在计算机视觉中的实现问题。

When dealing with complex scenes, biological vision can recognize the useful information fast and accurately. This is because that biological vision has the characteristic of selective attention. In the research domain of computer vision, much selective attention computing models have been proposed. This paper presented a survey on the selective attention computing models and algorithms. The research status was introduced. The deficiency of the current research was provided, toether with the future research directions.