SYSTEMATIC PSYCHOPHYSICAL INVESTIGATION OF VISUAL LEARNING

视觉感知学习（VPL）是指视觉任务表现的长期增强。VPL是一个 改善和/或恢复受损/退化视力的重要工具。VPL通常是专门为受过训练的 刺激。尽管VPL的特异性使人们对VPL的机制有了很大的了解，但它也被认为是一种新的研究方法。 严重限制了需要广泛效应的临床应用。在这份提案中，我们提出了 自然和日常场景的VPL对训练刺激的特异性低于 简单和不自然的刺激。在我们的日常生活中，视觉系统积极地与或接触到 由类别和其他特征组成的复杂的自然或日常场景。因此，在目标1和 第二，我们将首先使用简单但分类的方法来测试分类如何影响VPL的特异性程度， 刺激。在目标3中，我们将研究VPL是否以及如何由于暴露于NS而发生转移 图像. 具体目标1将考察类别学习对VPL迁移的影响（类别诱导迁移）。 VPL、CIT-VPL）。以下假设将通过心理物理学和脑成像进行检验。 假设（H）1-1a：CIT-VPL只发生在与受训者相同的类别中。 H1-1b：CIT-VPL越接近类别边界的方向越大。 H1-2：CIT-VPL是由训练类别（a）中未训练方向的性能提高引起的 或在未经训练的方向上表现下降，属于未经训练的类别（B）。 H1-3：CIT-VPL以特定位置的方式发生。 H1-4：早期视区在诱发CIT-VPL中起作用。 具体目标2将检查定向的VPL的CIT是否以及有多少被推广到定向的VPL的CIT。 运动方向 H2-1至H2-4将与H1-1至H1-4相同，除了运动方向的VPL的CIT，而不是 方向，将进行测试。 具体目标3将研究VPL的定向特异性如何受到统计和分类的影响， 自然场景。我们将从心理物理学的角度提出以下假设。 H3-1：由于暴露于NS而发生VPL。这一假设将通过VPL的方向进行检验， 受试者暴露于具有一定平均取向谱的任务无关的自然场景。 H3-2：与任务无关的NS的VPL比非自然刺激的VPL具有更小的位置特异性。我们将测试这个 通过检查视野一侧的自然景物的VPL是否转移到另一侧的假设 的方面想 H3-3：任务相关刺激的VPL受作为任务相关的背景NS类别的影响。

Visual perceptual learning (VPL) refers to a long-term enhancement of visual task performance. VPL is an important tool to improve and/or restore damaged/ degraded vision. VPL is usually specific for the trained stimulus. Although specificity in VPL has given great insights into the mechanisms of VPL, it is also regarded as a serious limitation to clinical applications that require generalized effects. In this proposal, we raise the possibility that VPL of natural and everyday scenes is less specific for the trained stimulus than VPL of a simple and unnatural stimulus. In our daily life, the visual system actively interact with or is exposed to complex natural or everyday scenes composed of categories and other characteristics. Thus, in Aims 1 and 2, we will first test how categorization influences the degree of specificity in VPL using simple but categorized stimuli. In Aim 3, we will examine whether and how transfer of VPL occurs as a result of being exposed to NS images. Specific Aim 1 will examine the effect of category learning on transfer of VPL (category-induced transfer of VPL, CIT-VPL). The following hypotheses will be tested by psychophysics and brain imaging. Hypothesis (H) 1-1a: CIT-VPL occurs only within the same category as the trained. H1-1b: CIT-VPL is greater with orientations closer to the category boundary. H1-2: CIT-VPL results from performance increases on untrained orientations within the trained category (a) or performance decreases on untrained orientations an untrained category (b). H1-3: CIT-VPL occurs in a location-specific manner . H1-4: Early visual areas play a role in inducing CIT-VPL. Specific Aim 2 will examine whether and how much CIT of VPL of orientation is generalized to CIT of VPL of motion direction. H2-1 to H2-4 will be identical to H1-1 to H1-4, except that CIT of VPL of motion direction, instead of orientation, will be tested. Specific Aim 3 will examine how orientation specificity of VPL is influenced by statistics and categories of natural scenes. We will the following hypotheses by psychophysics. H3-1: VPL occurs as a result of exposure to NSs. This hypothesis will be tested by VPL of orientations after subjects are exposed to task-irrelevant natural scenes with a certain mean orientation spectrums. H3-2: VPL of NSs as task-irrelevant is less location specific than VPL of unnatural stimuli. We will test this hypothesis by examining whether VPL of natural scenes on one side of the visual field transfers to the other side. H3-3: VPL of a task-relevant stimulus is influenced by the category of a background NS as task-relevant.