CENTRAL PROCESSING OF VISUAL INFORMATION

视觉信息的中央处理

• 批准号: 2161791
• 负责人: Jonathan D Victor
• 金额: $ 27.7万
• 依托单位: WEILL MEDICAL COLL OF CORNELL UNIV
• 依托单位国家: 美国
• 财政年份: 1989
• 资助国家: 美国
• 起止时间: 1989-01-01 至 1996-12-31
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/2161791
• 关键词: brain electrical activity; color visions; computational neuroscience; electroencephalography; evoked potentials; form /pattern perception; human subject; motion perception; neural information processing; psychophysics; visual feedback; visual pathways; visual perception; visual stimulus

A thorough understanding of visual information processing is important not only because vision is the largest source of sensory input to the nervous system, but also because the visual system is a model for neural information processing in general. The proposed research continues our attempts to advance the understanding of complex visual processing and how the submodalities of vision interact through the use of visual evoked potentials and psychophysics. In project [1], the neural computations underlying the extraction of visual form will be studied. The working model for this process consists of two levels of nonlinear processing: one, a local subunit which rectifies a high-pass transformation of the visual input, and a second longer-range nonlinearity which combines the output of these subunits in a supra-additive fashion. It is hypothesized that the-second nonlinearity adjusts its behavior in a manner dependent on local contrast. A main focus of the proposed work will be on how this neural measure of contrast is generated and how it tunes the second nonlinearity. In project [2], the interaction of color and luminance signals will be studied. A particular focus will be whether the roles of color and luminance signals in form analysis are a direct consequence of the way in which chromatic and luminance pathways (viewed as linear filters) transform visual information, or whether more elaborate models are required. This analysis may also provide clues as to how chromatic signals present at retinal and geniculate levels are transformed into the opponent colors of psychophysics. In project [3], the neural computations underlying the extraction of motion will be studied. The investigation will be organized around the hypothesis that the model framework of [1], along with appropriate time-delays, accounts for the extraction of motion from standard stimuli as well as "non-Fourier" motion stimuli. All of these studies are based on novel and sophisticated visual stimuli and analytical techniques; further development of such techniques is implicit in the proposed research.

对视觉信息处理的透彻理解很重要 不仅是因为视觉是最大的感官输入来源 神经系统，还因为视觉系统是神经的模型 一般信息处理。 拟议的研究继续我们 试图提高对复杂视觉处理和的理解 视觉唤起的使用方式如何相互作用 潜力和心理物理学。 在项目[1]中，提取的神经计算 将研究视觉形式。 此过程的工作模型包括 在两个非线性处理的两个级别中：一个，一个本地亚基 纠正视觉输入的高通转换，第二个 长期非线性结合了这些亚基的输出 超级添加的时尚。 假设这秒 非线性以取决于本地的方式调整其行为 对比。 拟议工作的主要重点将是这种神经如何 产生对比度的度量以及如何调整第二个 非线性。 在项目[2]中，颜色和亮度信号的相互作用将为 研究。 一个特别的重点将是颜色和 形式分析中的亮度信号是道路的直接结果 哪种色和亮度途径（被视为线性过滤器） 转换视觉信息，或者是否更精致的模型是 必需的。 该分析还可能提供有关色彩如何的线索 在视网膜和统一水平上存在的信号转化为 对手心理物理学的颜色。 在项目[3]中，提取的神经计算 将研究运动。 调查将围绕 假设[1]的模型框架以及适当的 时间纪念日，说明运动从标准刺激中提取 以及“非呼吸”运动刺激。 所有这些研究都是基于新颖而复杂的视觉刺激 和分析技术；这种技术的进一步发展是 拟议的研究隐含。