The time-course of human visual selection and recognition

在医学图像中寻找肿瘤是视觉注意力研究中经常提到的应用。使用各种复杂的仪表板在崎岖或危险的地形中驾驶车辆是另一种可以使用人类选择性注意力的应用知识来改善的情况。正是这些任务的共同性质，以及如果进一步理解和考虑人类视觉选择的特性，可以避免的潜在危险，使得对这些问题的讨论如此普遍。鉴于视觉可获得的信息量巨大，生物体的生存取决于快速选择这些信息的子集进行深入处理。 由于我们的感知和认知资源是有限的，我们的视觉系统必须迅速确定哪些部分的复杂场景是足够重要的，值得所需的资源进行充分的处理和适当的反应。早期选择机制影响几乎所有的视觉处理，这里提出的研究将增加我们的理解的方式相关的机制的行为和相互作用。此外，这些基本原则应该应用于改进任何需要人类用户注意或发现某些东西的设备的设计（例如，设计汽车仪表板以提高安全性，设计雷达屏幕以减少目标探测的失误）。在某些情况下，为人类注意力系统设计得不好的工具会产生比效率低下更严重的影响，例如驾驶员和飞行员的显示器。因此，除了澄清基本问题，并在理论领域作出贡献，这项工作将提供进一步的重要设计原则，有效的软件工具和用户界面设备的发展。

Finding tumours in medical images is a commonly mentioned application for research on visual attention. Navigating a vehicle through rugged or dangerous terrain, using a variety of complicated instrument panels, is another situation that could be improved using applied knowledge about human selective attention. It is the common nature of these tasks, and the potential dangers that could be avoided if the properties of human visual selection were further understood and taken into account, that have made discussions of these issues so ubiquitous. Given the overwhelming amount of information available to vision, an organism's survival depends on fast selection of subsets of that information for in-depth processing.  Because our perceptual and cognitive resources are limited, our visual systems must quickly determine what portions of a complex scene are important enough to merit the resources required for full processing and appropriate response.Early selection mechanisms affect virtually all visual processing, and the research proposed here will increase our understanding of the way the relevant mechanisms behave and interact. In addition, these basic principles should be applied to improving the design of any piece of equipment which requires a human user to notice or find something (e.g., design of car dashboards to increase safety, design of radar screens to decrease lapses in target detection). In some instances tools that are badly designed for the human attention system will have effects more drastic than mere inefficiency, for example displays for drivers and pilots. Thus, in addition to clarifying basic issues, and making a contribution in the theoretical domain, this work will provide further important design principles for the development of effective software tools and user interface devices.