Visual Coding and the Deployment of Attention

描述（由申请人提供）：本提案旨在继续和扩展我之前的研究，即不同的感知加工模式及其对不同分配注意力方式的依赖。注意力可以设置为选择环境的不同方面，从场景属性的全局注册到区分单个对象的局部特征连接的集中分析。在全局处理模式下，对大小判断的研究将包括对均值的内隐启动和对总体的虚幻表征的测试。将测试推广到其他维度，如颜色和方向，以及这种全球处理模式的各种容量限制。特征绑定中的集中注意力将通过改变显示的结构组织来测试，并测试在搜索中使用预先信息。一些fMRI和TMS研究也被提出来探索参与绑定任务不同方面的脑区，包括区分空间扫描、抑制干扰物和特征整合。还提出了“再入”账户约束的检验。第三个系列的实验探讨了视觉工作记忆中绑定的维持。提出了对象表征消失后短时保留的实验，比较了特征及其绑定的保留，并探索了容量限制和对干扰的脆弱性。其他研究将探索长期学习和视觉记忆之间的关系，看看我们暂时确定的短期和长期结合之间的分离是否以及以何种方式在其他测试中成立。最后一个项目分析了在对象的不同属性之间切换注意力时控制的不同组成部分——这可以被称为“解绑定问题”。它将使用快速分类范式来识别干扰源和控制组件，并通过功能磁共振成像在大脑中定位它们。对视觉注意力和记忆问题的回答应该有助于确定许多现实世界任务中潜在性能的系统架构，它们也可能提供一些理解，说明它们是如何在超负荷的情况下在正常人身上失灵的，或者在脑损伤导致忽视或巴林氏综合征的患者身上失灵的。

DESCRIPTION (provided by applicant): This proposal is to continue and extend the research begun on my previous grant on different modes of perceptual processing and their dependence on different ways of allocating attention. Attention can be set to select different aspects of the environment, ranging from the global registration of scene properties to the focused analysis of the local conjunctions of features that distinguish individual objects. In the global processing mode, studies of size judgements will include tests for implicit priming of the mean and for illusory representations of the population. Generalization to other dimensions such as color and orientation will be tested, as will various capacity limits inthis global processing mode. Focused attention in feature binding will be tested by varying the structural organization of the display, and testing the use of advance information in search. Some fMRI and TMS studies are also proposed to explore the brain areas involved in different aspects of the binding task, distinguishing spatial scanning, suppression of distractors and feature integration. Tests of a "re-entry" account of binding are also proposed. A third series of experiments explores the maintenance of binding in visual working memory. Experiments are proposed on the retention of object representations for short periods after they disappear, comparing the retention of features and of their bindings and exploring capacity limits and vulnerability to interference. Other studies will explore the relation between long term learning and visual memory to see whether and in what ways the dissociation between short and long term binding that we tentatively identified holds up in other tests. The last project analyses different components of control in switching attention between different attributes of an object - what could be called an "unbinding problem". It will use speeded classification paradigms to identify sources of interference and the components of control and to locate them in the brain through fMRI. Replies to the questions about visual attention and memory should help to determine the architecture of the systems underlying performance in many real world tasks, They may also provide some understanding of how they fail in normal people in conditions of overload, or in patients wtih brain injuries causing neglect or Balint's syndrome.