Simple Stimulus Metrics vs. Gestalt in High-Level Aftereffects
Simple Stimulus Metrics vs. Gestalt in High-Level Aftereffects
复制标题
简单刺激指标与高级后效中的格式塔
DOI：
10.1167/5.8.250
复制
发表时间：
2010
期刊：
影响因子：
--
通讯作者：
D. Leopold
中科院分区：
文献类型：
--
作者：
Kai;M. Ernst;D. Leopold
高水平的视觉后效（AE）出现时，适应刺激，如形状，面孔，或空间配置影响随后的感知可比的数字。与经典AE一样，高水平AE的特征在于在适应和测试刺激之间的向量的相反方向上在特征空间中被扭曲的感知。已报告了纵横比、凸度和锥度的形状对比不良事件，但这些参数的组合效应尚未探索。在本实验中，适应刺激由两条弧线组成。在保持弧的弯曲不变的同时，我们在±8.2的范围内改变它们之间的距离，使我们能够检查（1）纵横比和（2）凸度/曲率的影响。测试刺激是直径约7.8的封闭椭圆体。感知失真进行了评估与反复适应和测试的上下文中的楼梯程序收敛到主观循环点。我们发现这两个因素和相互作用的显着的主效应，即较小的纵横比导致更强的效果和凸刺激比凹的效果更大。适应和测试刺激的位置之间的直接空间对应对于诱发AE并不重要。我们用一组大约为1的较小刺激重复了这个实验，曲线之间的距离按比例缩小。两个实验的结果相当，尽管实验2中的相互作用不显著。结果表明，简单的刺激指标，如曲线的绝对大小和它们之间的距离，在创建此AE时不如由一对弧一起创建的整体形状重要。
High-level visual aftereffects (AEs) arise when adaptation to stimuli such as shapes, faces, or spatial configurations affect the subsequent perception of comparable figures. Like classical AEs, high-level AEs are characterized by a percept that is distorted in feature space in the opposite direction of the vector between adaptation and test stimulus. Shape-contrast AEs have been reported for aspect ratio, convexity and taper, but the combined effect of such parameters remained unexplored. In the present experiment the adapting stimulus consisted of two arcs. While keeping the flexion of the arcs constant we varied the distance between them in a range of±8.2, allowing us to examine the effects of (1) aspect ratio and (2) convexity/concavity. The test stimulus was a closed ellipsoid of about 7.8 in diameter. Perceptual distortions were assessed with repetitive adaptation and testing in the context of a staircase procedure converging to the point of subjective circularity. We found significant main effects for both factors and the interaction, ie smaller aspect ratios lead to stronger effects and convex stimuli result in larger effects than concave ones. The direct spatial correspondence between the location of the adapting and test stimulus was not critical for inducing an AE. We repeated the experiment with a set of smaller stimuli of roughly 1 and the distance between the curves scaled down proportionately. Results in both experiments were comparable, although the interaction was not significant in experiment 2. The results suggest that simple stimulus metrics, such as the absolute size of the curves and distance between them, are less important in creating this AE than the overall shape created by the pair of arcs together.