Controlling visual cognition with visual working memory and long-term memory

用视觉工作记忆和长期记忆控制视觉认知

• 批准号: 9247953
• 负责人: Gordon Dennis Logan
• 金额: $ 35.31万
• 依托单位: VANDERBILT UNIVERSITY
• 依托单位国家: 美国
• 财政年份: 2015
• 资助国家: 美国
• 起止时间: 2015-04-01 至 2019-03-31
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/9247953
• 关键词: Accounting; Address; Attention; Automobile Driving; Brain; Clinical; Clutterings; Cognitive; Complex; Computer Simulation; Costs and Benefits; Data; Digit structure; Electrophysiology (science); Environment; Foundations; Goals; Hand; ITAM; Learning; Letters; Maintenance; Measures; Memory; Methods; Modeling; Nature; Performance; Persons; Phase; Population; Research; Rewards; Role; Schools; Short-Term Memory; Source; Speed; Structure; Task Performances; Testing; Time; Ursidae Family; Visual; Visual attention; Visual system structure; attentional control; brain electrical activity; cognitive control; data modeling; indexing; information processing; long term memory; memory acquisition; novel; offspring; public health relevance; relating to nervous system; skill acquisition; theories; tool; visual cognition; visual control; visual learning; visual processing

 DESCRIPTION (provided by applicant): Theories of attention and learning differ in the roles they prescribe to memory representations. Theories of attention almost universally propose that visual working memory representations determine which objects are the focuses of processing in complex visual scenes. However, models of learning propose that after a handful of trials, the control of visual cognition should shift from being driven by visual working memory to representations in long-term memory. In this project we will bring these ideas from models of learning to bear on the processing of visual information. To test these model predictions in a definitive manner, we will develop methods of measuring electrical brain activity that we can use to track what information is being actively held in visual working memory, simultaneously with independent measures of the representations in visual long-term memory. Our preliminary data indicate that these electrophysiological tools can provide trial-by-trial indices of memory representations controlling visual cognition. Our integrated modeling and neuroscientific approach will allow us to more quickly lay a theoretical foundation for understanding the top-down control of visual processing. This can then be used to understand the nature of deficits in clinical populations and to refine potential treatments.

 描述（由应用程序提供）：注意力和学习的理论在他们准备为内存表示的角色中不同。注意力的理论几乎普遍提出，视觉工作记忆表示确定了哪些对象是在复杂的视觉场景中处理的重点。但是，学习建议的模型是，经过几次试验，视觉认知的控制应从视觉工作记忆驱动到长期记忆中的表示形式。在这个项目中，我们将从学习模型中带来这些想法，以遵循视觉信息的处理。为了以明确的方式测试这些模型预测，我们将开发用于测量电动脑活动的方法，我们可以使用这些方法来跟踪视觉工作记忆中正在积极保存的信息，仅使用视觉长期记忆中的表示形式独立衡量。我们的初步数据表明，这些电生理工具可以提供控制视觉认知的记忆表示的试用指标。我们集成的建模和神经科学方法将使我们更快地为理解自上而下的视觉处理控制奠定理论基础。然后，可以用来了解临床人群中缺陷的性质并完善潜在的治疗方法。