Abstract:

Object recognition has attracted a great deal of theoretical attention, and a number of competing explanations have been offered.The object-centered theory assumed an explicit treatment of structure in recognition, pointing out that the latter task can be made especially easy by the availability of a “structural description” of the object. By contrast, the view-centered theory posited that objects are represented as collections of entire viewpoint. In the current study, made 3D objects as experimental materials according to the idea of Biederman’s small geons; used the priming paradigm and a 4-alternative forced choice design. The participant responded to each target object by pressing one of four keys on the computer keyboard (‘F’ ,‘G’,‘H’,‘J’). The experiment 1 was conducted with a 3(displacement level on the two parts: no displacement, half displacement and full displacement) ×2(prime type: part and whole)within-subject design. The relative position of parts was manipulated by displacing one relative to the other. And the experiment 2 was conducted with a 3(location prime: same position, close position, distance position) ×2（prime type: part and whole）within-subject design. The relative location of objects was manipulated by displacing one relative to the other. The results are as follows: (1)the influence of displacement level and location is hierarchical. (2) On the no displacement and same location level , the speed of the whole prime was faster than the part prime. But on the full displacement and distance location level , the part prime was faster than the whole prime. These conclusions suppose that the “what” representation stage can be divided into part features and whole features, also the “where” representation stage can be divided into self-construction (displacement) and relative-construction(location).The finding not only extended the field of object recognition, but also shed light on the integration between the two object theories.

Key words: displacement, location, three-dimensional object, the object-centered theory, view-centered theory

摘要：

物体识别的两大理论一直存在争议。以物体为中心理论认为不管物体出现在什么位置，识别均与空间位置无关，而以观察者为中心理论认为识别与空间位置有关。研究参照物体识别的“小几何体”思想自制实验材料，采用启动范式下的分类任务，通过操纵物体自身的结构信息和相对的结构信息，考察了三维物体识别的影响机制。结果发现：（1）物体自身组成部分之间的分离水平和物体之间的相对空间位置对物体识别的影响均呈层级式。支持以观察者为中心理论的整体表征观；（2）不分离水平和相同位置上，整体启动快于部分启动；全分离水平和远距离位置上，部分启动快于整体启动。支持以物体为中心理论的小几何体优先加工观。实现两大理论的融合需要进一步厘清“What + Where”两通路联合表征的二级子层级。

关键词: 分离水平, 空间位置, 三维物体, 以物体为中心理论, 以观察者为中心理论