PROCESSING STREAMS IN EARLY VISION

描述（研究者摘要）：最近的证据表明， 视觉处理的早期阶段可以分为几个平行的阶段， 流，专门用于分析不同的视觉 美德.先知-愿 这里提出的研究将调查路由 信息通过这些早期的处理流使用五个刺激 媒体：亮度，颜色，纹理，立体声和运动。 特别是 研究人员感兴趣的是这些早期的相对贡献 流到刺激形状特征的最终确定（具体地 尺寸和方向），运动和深度，以及是否测量 形状，运动和立体声出现了几次（在每个 路径）或在共同表示中仅一次。 相反的刺激， 归纳和适应范式将被用来评估的程度， 每种介质分析的独立性。 初步实验 表明实际上重复了对二维形状特征类似分析 对于每一种媒介。 这种重复分析的一个原因可能是， 从多个介质中汇集信息可以提高轮廓的准确性 局部化，特别是当亮度信息不可靠时。 一 第二个原因是，使用类似的代码或图像格式， 这将有助于他们以后的融合。 另 另一方面，另一个试点实验表明，多个，中等特异性 不对运动进行分析。 情况可能类似于 双眼视差分析。 如果一种介质独立于其他介质进行分析，比如说，方向， 调查人员还将寻找编码独立性， 介质 例如，如果颜色定义的边框的方向为 独立于在其他文件中定义的边界的方向进行分析 媒体，这可能是负责的机制是有选择性的， 任何颜色边界的方向，无论涉及哪种颜色， 它们可以是特定于颜色，但对特定颜色没有选择性 (no独立编码）。 或者，它们可能都是颜色选择性的 以及方向选择性（独立性）。 同样的反对 上述归纳和适应范例将用于评估 介质内不同值分析的独立性。 的 这些研究的结果将与一组相应的 视觉搜索连词的实验。 这些视觉搜索任务 将结合相同的值（例如颜色和方向，颜色和运动） 与在相反的诱导和适应实验中测试的结果相同。 的 预测是，在相反的适应实验中， 与连接任务中缓慢的连续搜索相关。 的 解释在于干扰物组的异质性。 除了正常观察者的心理物理任务外， 研究者建议测试已知损失的患者（特别是， 皮质色盲），以验证信息的路由。 哪里 在数据的支持下，研究者应将这些测试作为 测量不同视觉通路的相对功能， 诊断特定疾病。

DESCRIPTION (Investigator's Abstract): Recent evidence indicates that the early stages in visual processing may be broken into several parallel streams that are specialized for the analysis of different visual attributes. The research proposed here will investigate the routing of information through these early processing streams using five stimulus media: luminance, color, texture, stereo and motion. In particular, the investigator is interested in the relative contributions of these early streams to the final determination of stimulus shape features (specifically size and orientation), motion and depth, and whether the measurement of shape, motion, and stereo emerges several times (repeated within each pathway) or only once in a common representation. Opposed stimuli, induction and adaptation paradigms will be used to evaluate the degree of independence of the analyses for each medium. Preliminary experiments suggest that a similar analysis of 2-D shape features is actually repeated for each medium. One reason for this duplication of analyses may be that pooling information from several media can improve the accuracy of contour localization particularly when luminance information is unreliable. A second reason is that using similar codes or image formats for each representation will facilitate their subsequent integration. On the other hand, another pilot experiment suggests that multiple, medium-specific analyses do not occur for motion. The situation may be similar for analyses of binocular disparity. If a medium is analyzed independently of other media for, say, orientation, the investigator will also look for independence of coding within that medium. For example, if the orientation of color-defined borders is analyzed independently of the orientation of borders defined in other media, it may be that the mechanisms responsible are selective for the orientation of any color border no matter which colors are involved -- that is, they may be specific to color but not selective for particular colors (no independent coding). Alternatively, they might be both color-selective as well as orientation-selective (independence). The same opposing induction and adaptation paradigms mentioned above will be used to evaluate the independence of analysis for different values within a medium. The results of these studies will be compared to a set of corresponding experiments on visual search for conjunctions. These visual search tasks will conjoin the same values (e.g. color and orientation, color and motion) as those tested in the opposing induction and adaptation experiments. The prediction is that independence in the opposing adaptation experiments will correlate with slow, serial search in the conjunction task. The explanation lies in the heterogeneity of the distractor groups. In addition to psychophysical tasks with normal observers, the investigators propose to test patients with known losses (in particular, cortical color blindness) to verify the routing of information. Where supported by the data, the investigators shall extend these tests as a measure of relative function of different visual pathways that may be diagnostic of particular diseases.