Recognizing depth-rotated objects: A review of recent research and theory
Recognizing depth-rotated objects: A review of recent research and theory
复制标题
DOI：
10.1163/156856800741063
复制
发表时间：
2000-01-01
期刊：
影响因子：
--
通讯作者：
Biederman, I
中科院分区：
文献类型：
--
作者：
Biederman, I
对象分类的许多现象可以从指定对象的部分和关系的非偶然特征的表示中导出，称为对象结构描述（GSD）。这样的表示：(a)能够轻松识别深度旋转的对象，即使它们是新的，(b)提供的信息不仅用于区分基本级别，而且还用于区分从属级别类的高度相似的成员，(c)能够映射到口头和对象推理结构。最近对物体识别的心理物理和神经研究为这种物体表征理论提供了额外的支持。
Many of the phenomena of object classification can be derived from a representation specifying a nonaccidental characterization of an object's parts (geons) and relations, termed a geon structural description (GSD). Such a representation: (a) enables the facile recognition of depth-rotated objects, even when they are novel, (b) provides the information that is employed not only to distinguish basic-level but also highly similar members of subordinate-level classes, and (c) enables mapping onto verbal and object-reasoning structures. Recent psychophysical and neural investigations of object recognition have provided additional support to this theory of object representation.