Construction of a database of shape patterns with psychological variables

(1)为了给形状感知和识别的研究提供一个标准化的刺激集，我们开发了一个无义闭合轮廓形状及其各种心理特性的数据库。在发展过程中，我们考虑了对心理学研究的有用性，与自然物体共享的共同属性，以及几何分析技术的可用性。在为本研究创建的许多无意义形状中，选择了100个形状进行分析。我们通过排序任务构建心理相似空间，并测量了各种心理属性，如复杂性和对称性。我们还进行了一项任务来测量无意义形状与已知自然物体的关联。相似性、复杂性、对称性等心理属性之间的关系分析；傅立叶描述子和圆度等几何性质揭示了复杂性和圆度之间的相关性，以及心理相似空间与傅立叶描述子导出的相似空间之间的对应关系。对于关联数据，我们提出了基于KL信息的关联传播指数，该指数能够衡量关联的可变性，并显示了关联传播指数的有效性。(2)对无意义形状的感知和记忆进行了心理实验。(a)在无意义形状的选择性注意任务中，与以往关于注意力对部分遮挡图形的影响的理论不同，我们证明了注意力的影响实际上可以用图形对称性的影响来解释。(b)利用无义轮廓形状研究了视觉短期记忆的时空特性。在需要存储精确形状的物体识别任务中，以凸部分的数量作为物体复杂程度的指标，物体凸部分越多，识别性能越差。对数据的详细分析表明，物体被凸部分分割，每个凸部分被感知并存储在记忆中。(c)为了探讨无义形状匹配任务中负启动的发生机制，目标选择的难度受双眼刺激深度的影响。结果表明，负启动的决定因素不是目标形状和干扰物形状之间的干扰，而是标准形状和干扰物形状之间的干扰。少

(1) To provide a standardized stimulus set for studies of shape perception and recognition, we have developed a database of nonsense closed contour shapes with their various psychological properties. In the development, we have considered the usefulness for psychological research, the common properties sharing with natural objects, and the availability of techniques of geometric analyses. Among the many nonsense shapes created for this study, one hundred shapes were selected for the analyses. We constructed psychological similarity spaces by a sorting task, and measured various psychological properties such as complexity and symmetry. We also conducted a task to measure associations of the nonsense shapes with known natural objects. The analyses of the relations between psychological properties such as similarity, complexity, and symmetry ; and geometric properties such as Fourier descriptors and roundedness revealed that correlation between complexity and roundedness, and the correspo … More ndence between the psychological similarity space and the similarity space derived from Fourier descriptors. As for the association data, we proposed an association-spreading index, which is based on KL information and is able to measure the variability of associations, and showed the efficiency of the association-spreading index.(2) We conducted psychological experiments on perception and memory of nonsense shapes. (a) Inconsistent with the previous theories on the effect of attention on partially occluded figures in a selective attention task with nonsense shapes, we have shown that the effects of attention in fact can be accounted for by the effect of symmetry of the figures. (b) We investigated the spatiotemporal properties of visual short-term memory using nonsense contour shapes. In a recognition task where participants need to store precise shape of objects and the number of convex parts was used as an index of the complexity of objects, the more an object had convex parts, the less the recognition performance was. Detailed analyses of the data suggest that objects were segmented by convex parts, and each convex part was perceived and stored in memory. (c) To investigate the mechanism of the occurrence of negative priming in a matching task of nonsense shapes, the difficulty of target selection was manipulated by the binocular depth of the stimuli. The results showed that the determinant of the occurrence of negative priming was not the interference between target and distracter shapes, but the interference between the standard and distracter shapes. Less